Synopsis: Ict:

Any comments can be sent by email to: JRC-IPTS-IRI@ec. europa. eu More information, including activities

+34 95 448 83 18, Fax:++34 95 448 83 00 IPTS e-mail: jrc-ipts-secretariat@ec. europa. eu Legal Notice:

Further information about the European union is available on the Internet. It can be accessed via the Europa web portal at:

http://europa. eu EUR 26224 EN ISBN 978-92-79-33748-2 (print), 978-92-79-33747-5 (pdf) ISSN

the overall expectations of all the other companies in the sample show a more positive outlook for industrial R&d at exactly the same global level as in past year's survey (4%).For some sectors,

%and technology hardware & equipment (4%).Figure 1: Expected changes of R&d investment of the surveyed companies 2013-15, p. a. Note:

Million euros in 2011.5%0%5%10%Software & Computer Services Pharmaceuticals & Biotechnology Technology Hardware & Equipment Health care Equipment & Services Electronic & Electrical Equipment General

while maintaining an R&d focus in the EU. Low expectations for R&d in the EU (1%p. a. in 2013-15) are due to the outlook of seven automobiles

Knowledge-sharing, human resources, proximity to other company sites and market demand make countries attractive for R&d activities.

the respondents state that knowledge-sharing and collaboration opportunities with universities and public research organisations, quality and quantity of R&d personnel in the labour market, proximity to other company sites,

Geographic proximity to other company sites is attractive for R&d in Germany and the UK,

The respondents considered the US a more attractive site for R&d activity than the EU especially in terms of market size and growth,

for the EU geographic proximity to other company sites and technology poles & incubators is a factor for attractiveness.

and monitors progress towards the 3%headline target. The survey complements IRIMA's core activity, the EU Industrial R&d Investment Scoreboard

7 which analyses private R&d investments based on the audited annual accounts of companies and shows ex-post trends.

and provides data and analysis on companies from the EU and abroad investing the largest sums in R&d (see:

Technology Hardware & Equipment, Software & Computer Services, and Health care Equipment & Services 49 47%Medium R&d intensity Industrial Engineering, Electronic & Electrical Equipment, Automobiles & Parts, Chemicals, Aerospace & Defence, General Industrials

, Household Goods & Home Construction, Food Producers, Travel & Leisure, Financial services, Fixed Line Telecommunications, Alternative energy, Support Services, Equity Investment Instruments,

Their outlook was compared significantly lower to the past(-0. 7%p. a. for 2013-15 vs. around 5%in our two previous surveys

%.While that level is a positive outlook for corporate R&d above the nominal EU GDP growth estimates at 1. 4%for 2013 and 1. 9%for 2014,15 the R&d investment expectations are not yet at the levels

In the high R&d intensity group, expected R&d investment changes from pharmaceuticals & biotechnology (4. 4%)and technology hardware & equipment (3. 6%)are slightly above those of last year's survey

-5%0%5%10%15%Software & Computer Services Pharmaceuticals & Biotechnology Technology Hardware & Equipment Health care Equipment & Services Electronic & Electrical Equipment General

their expected R&d investment changes are inline to the expected vehicle sales outlook for the next years,

& parts companies in China R&d investment of the 9 surveyed companies in China passenger vehicles sales outlook in China 14 The 2013 EU SURVEY on R&d Investment Business

also for US companies the 2013 outlook for R&d investment changes has been reduced to 2. 3%18 due to more moderate growth dynamics compared to the previous period. 19 The comparison of R&d investment

%In the high R&d intensity sectors, pharmaceuticals & biotechnology and software & computer services are the drivers of expectations in the US and Canada, China and India.

without the need to refer to actual R&d sites. In this context, two thirds of the respondents considered their home country as the most attractive location.

Figure 13 displays the ranking of the most attractive country for outsourcing the company's R&d to other companies.

This question allows for a pairwise comparison of the actual R&d sites. Similar to the observations above, nine out of ten respondents stated their home country as one of the two with the highest volume of R&d activity (Figure 14.

Above average attractiveness was stated for knowledge-sharing and collaboration opportunities with universities and public research organisations, quality and quantity of R&d personnel in the labour market, proximity to other company sites,

& public organisations with other firms quality quantity labour costs of R&d personnel other company sites technology poles

sites (Germany and the UK), and public R&d support via fiscal incentives (France and Spain) or once (IPR enforcement conditions (Belgium), proximity to suppliers (Spain) and labour costs of R&d personnel (Poland

with universities & public organisations proximity to other company sites public R&d support via fiscal incentives France (25) 3

knowledge-sharing opportunities with universities & public organisations quality of R&d personnel proximity to other company sites innovation demand via product market regulation Sweden (12

& incubators other company sites suppliers enforcement conditions time to obtain protection costs grants & direct funding fiscal incentives public-private partnerships loans & guarantees financing other investments market size via product market regulation market growth via public procurement Knowledge-sharing opportunities

The respondents considered the US a more attractive site for R&d activity than the EU especially in terms of market size and growth,

and US. 1 2 3 4 5 other company sites technology poles & incubators suppliers with universities & public organisations with other firms quality quantity

Proximity is on average the most important factor here, in the case for China and India related to suppliers and for the EU to other company sites and technology poles & incubators.

it should be emphasised that they correspond to actual cases of considerable R&d activity by leading companies in these countries. 1 2 3 4 5 suppliers other company sites technology poles

norms & standards costs of protection IPR conditions IPR time to obatain protection suppliers other company sites technology poles & incubators quantity of R&d personnel

%)Firms across all sector groups value the acquisition of new or highly improved machinery, equipment and software within the European union higher than acquisition from outside (non-EU) countries.

Companies in the technology hardware & equipment and pharmaceuticals & biotechnology (high R&d intensity) report the highest average shares.

and some occasional country-specific statistics, were the main sources of these data. 32 A mapping of available transnational data sources on industrial R&d33 from the European commission,

OECD and European industry associations, showed that data on business enterprise R&d essentially drew upon retrospective surveys

Statistical offices generally collect R&d data in the form of Business R&d Expenditure (BERD which defines R&d from a top-down perspective.

Private data sources and surveys by industrial associations existed but were published rarely, and there was a shortage of qualitative and forward-looking information on industrial R&d.

and policy making in this area was usually based on results of analysis based on partial or incomplete data.

The survey complements other R&d investment related surveys and data collection exercises (e g. Innobarometer, Eurostat data collection and other ongoing surveys.

Link to the R&d Investment Scoreboards The EU R&d survey is part of the Industrial Research

Mapping Surveys and other Data Sources on Industrial R&d in the EU-25 countries, Seville,

Description of Information Sources on Industrial R&d data: European commission, OECD and European Industry Associations, Seville, July 2004.34 The rationale for the IRIMA activities emerged in the context of the European commission's"3%Action Plan"established to implement

and provides data and analysis on the largest R&d investing companies in the EU and abroad (see:

To maintain the maximum information in the data, outliers were eliminated only in extreme cases and after assessing the impact on the result. 37 One-year growth is simple growth over the previous year,

only if data exist for both the current and previous year. At the aggregate level, 1yr growth is calculated only by aggregating those companies for

which data exist for both the current and previous year. Three-year growth is the compound annual growth over the previous three years,

only if data exist for the current and base years. At the aggregate level, 3yr growth is calculated only by aggregating those companies for

which data exist for the current and base years. Unless otherwise stated, the weighted figures presented in this report are weighted by R&d investment.

The 2013 questionnaire has a rather high number of items compared to its predecessors due to the coverage of country comparisons in questions 6 to 8. 2) The questionnaire was sent together with the Scoreboard report to take advantage of this occasion as a door-opener.

an online site was provided to facilitate data entry via the European commission's Interactive Policy-making (IPM) tool,

38 where a Word version of the questionnaire was downloadable for offline information input. 5) The questionnaire was emailed to the respondents of previous surveys,

7) The contact database was improved continuously. Respondents who had participated already in previous surveys, or their substitutes in cases where they had left their position,

Returned questionnaires and reminder mailings were resent using the latest contact information on the internet

or by contacting the company directly via email or phone. 8) The response rate is followed closely on a regular basis during the implementation.

allowing more time for questionnaire reception, following up selected candidates by e-mail and phone or searching support from former survey participants (9) Personal contact,

mostly by phone, was made with several dozen companies when the deadlines were close, especially for those

pharmaceuticals & biotechnology, technology hardware & equipment, software & computer services, health care equipment & services,

industrial engineering, chemicals, aerospace & defence, electronic & electrical equipment, automobiles & parts, general industrials, fixed line telecommunications, food producers, alternative energy, household goods

and mobile telecommunications. Table 3 shows the distribution of the responses among the sectors with their respective R&d investment shares. responses received per day of the response period has doubled almost,

sector group**Pharmaceuticals & Biotechnology 24 108 22.2%above 40%High technology Hardware & Equipment 10 47 21.3%above 40%High Software & Computer Services 8

This is the result of the high share of R&d employees in large companies that responded from technology, hardware & equipment and pharmaceuticals & biotechnology (high R&d intensity), automobiles & parts, industrial engineering,

c1) technology poles52 and incubators53 (c2) other company sites, e g. production or sales (c3) suppliers (d) Collaboration & knowledge-sharing opportunities:(

c1) Inside the European union (c2) In non-EU countries (d) Acquisition of new or highly improved machinery, equipment and software:(

The European union is committed to data protection and privacy as defined in Regulation (EC) n 45/2001.

The Controller commits himself dealing with the data collected with the necessary confidentiality and security as defined in the regulation on data protection and processes it only for the explicit and legitimate purposes declared

and will not further process it in a way incompatible with these purposes. These processing operations are subject to a Notification to the Data protection Officer (DPO) in accordance with Regulation (EC 45/2001.

Purpose and data treatment The purpose of data collection is to establish the analysis of the 2013 EU Survey of R&d Investment Business Trends.

This survey has a direct mandate from the Commission's 2003 Action Plan"Investing in Research"(COM 2003 (226) final,

see http://ec. europa. eu/invest-in-research/action/2003 actionplan en. htm). The personal data collected and further processed are:

e-mail The collected personal data and all information related to the above mentioned survey is stored on servers of the JRCIPTS, the operations

and provisions established by the Directorate of Security for these kind of servers and services. The information you provide will be treated as confidential

and aggregated for analysis. Data verification and modification In case you want to verify the personal data or to have modified it respectively corrected,

or deleted, please write an email message to the address mentioned under Contact information, by specifying your request.

Special attention is drawn to the consequences of a delete request, in which case any trace to be able to contact you will be lost.

Your personal data is stored as long as follow-up actions to the above mentioned survey are necessary with regard to the processing of personal data.

Contact information In case you have questions related to this survey, or concerning any information processed in this context,

operating under the responsibility of the Controller at the following email address: jrc-ipts-iri@ec. europa. eu. Recourse Complaints,

in case of conflict, can be addressed to the European Data protection Supervisor (EDPS) at www. edps. europa. eu. European commission EUR 26224 EN Joint Research Centre Institute for Prospective Technological Studies

%and technology hardware & equipment (4%).The responding companies carry out a quarter of their R&d outside the EU. Their expectations for R&d investment for the next three years show continued participation of European companies in the global economy, in particular growth

Knowledge-sharing, human resources, proximity to other company sites and market demand make countries attractive for R&d activities.

for the EU geographic proximity to other company sites and technology poles & incubators is a factor for attractiveness.

You can obtain their contact details by sending a fax to (352) 29 29-42758.

Afterexaminingandcleaning the data, 448firmswereusedinthisanalysis. In thisstudy, wedefinealistofpossiblefactorsthathave bearing oninnovation (Tables1 4). Ourgoalistofind those factorsthathavesignificantimpactoninnovationin SMES inasmalldevelopingcountry.

Data showthatthereisnodifferenceinprocess innovationbetweenfirmsthatreportobstaclesandthose that donot (N 172, w2 1. 9, p 0. 17. Regarding productinnovation, thereisaweakrelationshipshowing that farfrombeing less innovative, firmsthatreported obstaclesare more innovativecomparedwithotherfirms that didnotreportobstacles (81.16%ofthosethat reportedobstaclesinnovatedcomparedwith68. 93%of those thatdidnotreportobstacles N 172, w2 3. 2 and p

a multi-site casestudyoffamilyownedbusiness. Journalofbusinessand Entrepreneurship1 (2), 41 58. Hoffman, K.,Parejo, M.,Bessant, J.,Perren, L.,1998.

and ethnic distinctions) and a reflexive approach to knowledge and practices among the core competencies that are crucial in creating A Culture of Innovation.

With the increasing importance of Information and Communication Technologies,(ICTS), the digital divide has grown at a rapid pace.

such as the digital divide which increases the development gap, free circulation and equal access to data, information and to good practices and the knowledge of information societies,

Lecture Notes in Computer science 6656 Commenced Publication in 1973 Founding and Former Series Editors: Gerhard Goos, Juris Hartmanis,

Switzerland C. Pandu Rangan Indian Institute of technology, Madras, India Bernhard Steffen TU Dortmund University, Germany Madhu Sudan Microsoft Research, Cambridge, MA

The Future Internet Future Internet Assembly 2011: Achievements and Technological Promises 13 Volume Editors John Domingue Alex Galis Anastasius Gavras Theodore Zahariadis Dave Lambert Frances Cleary Petros Daras

SL 5 Computer Communication Networks and Telecommunications The Editor (s)( if applicable) and the Author (s) 2011.

and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version,

Information systems, University of Applied sciences Western Switzerland, Sierre, Switzerland henning. mueller@hevs. ch VI List of Editors Man-Sze Li IC Focus, London

Sweden michael. nilsson@cdt. ltu. se Foreword The Internet will be a catalyst for much of our innovation and prosperity in the future.

A competitive Europe will require Internet connectivity and services beyond the capabilities offered by current technologies.

Future Internet research is therefore a must. Since the signing of the Bled declaration in 2008,

European research projects are developing new technologies that can be used for the Internet of the Future.

At the moment around 128 ongoing projects are being conducted in the field of networks, trustworthy ICT, Future Internet research and experimentation,

services and cloud computing, networked media and Internet of things. In total they represent an investment in research of almost 870 million euro,

as these projects meet twice a year during the Future Internet Assembly, where they discuss research issues covering several of the domains mentioned above,

Apart from the Future Internet Assembly, the European commission has launched also a Public Private Partnership program on the Future Internet.

The core of this program will be a platform that implements and integrates new generic but fundamental capabilities of the Future Internet,

such as interactions with the real world through sensor/actuator networks, network virtualization and cloud computing, enhanced privacy and security features and advanced multimedia capabilities.

This core platform will be based on integration of already existing research results developed over the past few years,

and will be tested on large-scale use cases. The use cases that are part of the Public Private Partnership all have the potential to optimize large-scale business processes,

using the properties of the core Future Internet platform. Examples of these use cases are a smarter electricity grid, a more efficient international logistics chain

Future Internet research is an important cornerstone for a competitive Europe. We believe that all these efforts will help European organizations to be in the driving seat of many developments of the Future Internet.

This book, already the third in this series, presents some of the results of this endeavor.

The uniqueness of this book lies in the breadth of the topics, all of them of crucial importance for the Future Internet.

VIII Foreword We sincerely hope that reading it will provide you with a broader view on the Future Internet efforts and achievements in Europe!

Budapest, May 2011 Luis Rodríguez-Roselló Mário Campolargo Preface 1 The Internet Today Whether we use economic or societal metrics,

the Internet is one of the most important technical infrastructures in existence today. One easy measure of the Internet's impact and importance is the number of Internet users

which as of June 2010 was 2 billion1. But of course, this does not give one the full picture.

From an economic viewpoint, in 2010 the revenue of Internet companies in the US alone was over $70 billion2.

In Europe, IDC estimated that in 2009 the broader Internet revenues (taking business usage into account) amounted to €159 billion

The recent political protests in Egypt give us an indication of the impact the Internet has in societal terms.

At the start of the demonstrations in Egypt the Internet was closed down by the ruling government to hinder the activities of opposition groups.

Protesters in Egypt used social media to support communication and the associated Facebook page had over 80,000 followers at its peak.

It is interesting to note that here we are talking about the power of the Internet in a country where currently Internet penetration is compared 21%5 to say 79%for Germany6. 2 Current Issues The Internet has recently been in the news with stories covering two main issues

which are known commonly in the Internet research community. Firstly, recent stories have highlighted the issue of the lack of address space associated with IPV4,

which can cater for 4 billion IP addresses7. Some headlines claim that the IPV4 address space has already run out8.

Technically, the issue has been solved through IPV6 although there is still the matter of encouraging take up. 1 http://www. internetworldstats. com/stats. htm 2 http://money. cnn. com/magazines/fortune/fortune500/2010/industries/225/index

. html 3 http://www. fi3p. eu 4 http://www. mediaite. com/tv/picture-of-the-day-cairo-protester-holds-sign-that-saysthank-you-facebook/5

http://www. internetworldstats. com/africa. htm#eg 6 http://www. internetworldstats. com/europa. htm#de 7 http://www. bbc. co

. uk/news/10105978 8 http://www. ndtv. com/article/technology/internet-will-run out-of-ip-addresses-by-friday-83244 X Preface A second major news item has been on net

neutrality, specifically, on legislation on net neutrality in the US and UK, which take differing views.

Other issues are centered on the fact that the Internet was designed originally in a very different context

Of the changes that have occurred in the decades since the Internet's inception the main alterations which are of concern are:

Volume and nature of data the sheer volume of Internet traffic and the change from simple text characters to audio and video and also the demand for very immediate responses.

For example, Cisco's latest forecast predicts that global data traffic on the Internet will exceed 767 Exabytes by 2014.

Mobile devices the Internet can now be accessed from a wide variety of mobile devices including smart phones

Internet radios, and vehicle navigation systems, which is a radically different environment from the initial Internet based on physical links.

Data traffic for mobile broadband will double every year until 2014, increasing 39 times between 2009 and 201413.

Physical objects on the net small devices enable the emergence of the Internet of things where practically any physical object can now be on the net sending location and local context data when requested.

Commercial services as mentioned above the Internet is now a conduit for a wide variety of commercial services.

These business services rely on platforms which can support a wide variety of business transactions and business processes.

The general population demand that the Internet is at least: secure, trustworthy, ubiquitous, robust, responsive and also upholds privacy. 9 http://online. wsj. com/article/BT-CO-20110217-718244. html 10 http://www. bbc. co. uk/news

id=long-live-the-web, http://www. theatlantic. com/technology/archive/2010/12/steve-wozniak-to-the-fcc-keep-theinternet-free/68294/12 http://www. ispreview. co. uk/story

-jump-in-globalinternet-traffic-by-2014. html Preface XI 3 FIA Overview This book is based on the research that is carried out within the Future Internet Assembly (FIA).

In short, FIAS bring together over 150 research projects that are part of the FP7 Challenge 1 ICT Programme to strengthen Europe's Future Internet research activities

The network of the future Cloud computing, Internet of services and advanced software engineering Internet-connected objects Trustworthy ICT Networked media

and search systems Socioeconomic considerations for the Future Internet Application domains for the Future Internet Future Internet research and experimentation (FIRE) Researchers

and practitioners associated with the Future Internet gather at the FIAS every six months for a dialogue and interaction on topics

In conjunction with the meetings the FIA Working groups sustain activity throughout the year working toward a common vision for the Future Internet based on scenarios and roadmaps.

An overview of FIAS and the FIA working groups can be found at the EU Future Internet portal:

In the middle of 2010 a call was issued for abstracts of up to 2 pages covering a relevant Future Internet topic.

Foundations-Architectural Issues-Socioeconomic Issues-Security and Trust-Experiments and Experimental Design Future Internet Areas-Networks-Services-Content Applications FIA Budapest will be the seventh FIA

Future Internet Foundations: Architectural Issues Introduction to Part I...3 Towards a Future Internet Architecture...

7 Theodore Zahariadis, Dimitri Papadimitriou, Hannes Tschofenig, Stephan Haller, Petros Daras, George D. Stamoulis, and Manfred Hauswirth Towards In-Network Clouds in Future Internet...

19 Alex Galis, Stuart Clayman, Laurent Lefevre, Andreas Fischer, Hermann de Meer, Javier Rubio-Loyola, Joan Serrat,

Towards Scalable Future Internet Mobility...35 L'aszl'o Bokor, Zolt'an Faigl, and S'andor Imre Review and Designs of Federated Management in Future Internet Architectures. 51 Mart'in Serrano, Steven Davy, Martin Johnsson, Willie Donnelly,

and Alex Galis An Architectural Blueprint for a Real-world Internet...67 Alex Gluhak, Manfred Hauswirth, Srdjan Krco, Nenad Stojanovic, Martin Bauer, Rasmus Nielsen, Stephan Haller, Neeli Prasad, Vinny Reynolds,

and Oscar Corcho Towards a RESTFUL Architecture for Managing a Global Distributed Interlinked Data-Content-Information Space...

81 Maria Chiara Pettenati, Lucia Ciofi, Franco Pirri, and Dino Giuli A Cognitive Future Internet Architecture...

91 Marco Castrucci, Francesco Delli Priscoli, Antonio Pietrabissa, and Vincenzo Suraci Title Model Ontology for Future Internet Networks...

103 Joao Henrique de Souza Pereira, Flavio de Oliveira Silva, Edmo Lopes Filho, Sergio Takeo Kofuji,

Future Internet Foundations: Socioeconomic Issues Introduction to Part II...117 XIV Table of contents Assessment of Economic Management of Overlay Traffic:

and Burkhard Stiller Deployment and Adoption of Future Internet Protocols...133 Philip Eardley, Michalis Kanakakis, Alexandros Kostopoulos, Tapio Lev a, Ken Richardson,

and Henna Warma An Approach to Investigating Socioeconomic Tussles Arising from Building the Future Internet...

Future Internet Foundations: Security and Trust Introduction to Part III...163 Security Design for an Inter-Domain Publish/Subscribe Architecture...

and Sasu Tarkoma Engineering Secure Future Internet Services...177 Wouter Joosen, Javier Lopez, Fabio Martinelli,

and Fabio Massacci Towards Formal Validation of Trust and Security in the Internet of Services...

and Luca Vigan`o Trustworthy Clouds Underpinning the Future Internet...209 R udiger Glott, Elmar Husmann, Ahmad-Reza Sadeghi,

and Matthias Schunter Data Usage Control in the future Internet Cloud...223 Michele Bezzi and Slim Trabelsi Part IV:

Future Internet Foundations: Experiments and Experimental Design Introduction to Part IV...235 A Use-Case on Testing Adaptive Admission Control and Resource Allocation Algorithms on the Federated Environment of Panlab...

237 Christos Tranoris, Pierpaolo Giacomin, and Spyros Denazis Multipath Routing Slice Experiments in Federated Testbeds...

and Carsten Schmoll Table of contents XV Testing End-to-end Self management in a Wireless Future Internet Environment 259 Apostolos Kousaridas George Katsikas, Nancy Alonistioti, Esa Piri, Marko Palola,

Future Internet Areas: Network Introduction to Part V...273 Challenges for Enhanced Network Self-Manageability in the Scope of Future Internet Development...

277 Ioannis P. Chochliouros, Anastasia S. Spiliopoulou, and Nancy Alonistioti Efficient Opportunistic Network Creation in the Context of Future Internet...

293 Andreas Georgakopoulos, Kostas Tsagkaris, Vera Stavroulaki, and Panagiotis Demestichas Bringing Optical Networks to the Cloud:

An Architecture for a Sustainable Future Internet...307 Pascale Vicat-Blanc, Sergi Figuerola, Xiaomin Chen, Giada Landi, Eduard Escalona, Chris Develder, Anna Tzanakaki, Yuri Demchenko, Joan A. Garc

Future Internet Areas: Services Introduction to Part VI...323 SLAS Empowering Services in the future Internet...327 Joe Butler, Juan Lambea, Michael Nolan, Wolfgang Theilmann, Francesco Torelli, Ramin Yahyapour, Annamaria Chiasera,

and Marco Pistore Meeting Services and Networks in the future Internet...339 Eduardo Santos, Fabiola Pereira, Jo ao Henrique Pereira, Luiz Cl'audio Theodoro, Pedro Rosa,

and Sergio Takeo Kofuji Fostering a Relationship between Linked Data and the Internet of Services...

351 John Domingue, Carlos Pedrinaci, Maria Maleshkova, Barry Norton, and Reto Krummenacher Part VII: Future Internet Areas:

Content Introduction to Part VII...367 XVI Table of contents Media Ecosystems: A Novel Approach for Content-Awareness in Future Networks...

and C. Timmerer Scalable and Adaptable Media Coding Techniques for Future Internet...381 Naeem Ramzan and Ebroul Izquierdo Semantic Context Inference in Multimedia Search...

Future Internet Applications Introduction to Part VIII...403 Future Internet Enterprise Systems: A Flexible Architectural Approach for Innovation...

407 Daniela Angelucci, Michele Missikoff, and Francesco Taglino Renewable Energy Provisioning for ICT Services in a Future Internet...

419 Kim Khoa Nguyen, Mohamed Cheriet, Mathieu Lemay, Bill St. Arnaud, Victor Reijs, Andrew Mackarel, Pau Minoves, Alin Pastrama,

and Ward Van Heddeghem Smart Cities and the Future Internet: Towards Cooperation Frameworks for Open Innovation...

and Alvaro Oliveira Smart Cities at the Forefront of the Future Internet...447 Jos'e M. Hern'andez-Mu noz, Jes'us Bernat Vercher, Luis Mu noz, Jos'e A. Galache, Mirko Presser, Luis

Future Internet Foundations: Architectural Issues Part I: Future Internet Foundations: Architectural Issues 3 Introduction The Internet has evolved from a slow, person-to-machine, communication channel to the most important medium for information exchange.

Billions of people all over the world use the Internet for finding, accessing and exchanging information,

enjoying multimedia communications, taking advantage of advanced software services, buying and selling, keeping in touch with family and friends,

to name a few. The success of the Internet has created even higher hopes and expectations for new applications and services,

which the current Internet may not be able to support to a sufficient level. On one hand the increased reliability, availability and interoperability requirements of the new networked services,

and on the other hand the extremely high volumes of multimedia content challenge the today's Internet. As a result, the Future Internet research and development threads have been gaining momentum all over the world

and as such the international race to create a new generation Internet is in full swing. The current Internet has been founded on a basic architectural premise, that is:

a simple network service can be used as a universal means to interconnect both dumb and intelligent end systems.

The simplicity of the current Internet has pushed complexity into the endpoints, and has allowed impressive scale in terms of interconnected devices.

However, while the scale has reached not yet its limits, the growth of functionality and the growth of size have slowed both down

and may soon reach both its architectural capability and capacity limits. The current Internet capability limit will be stressed further by the expected growth, in the next years, in order of magnitude of more Internet services

the likely increase in the interconnection of smart objects and items (Internet of things) and its integration with enterprise applications.

Although the current Internet, as a ubiquitous and universal means for communication and computation, has been extraordinarily successful,

there are still many unsolved problems and challenges some of which have basic aspects. Many of these aspects could not have been foreseen

when the first parts of the Internet were built, but these do need to be addressed now. The very success of the Internet is now creating obstacles to the future innovation of both the networking technology that lies at the Internet's core and the services that use it.

We are faced with an Internet that is good at delivering packets but shows a level of inflexibility at the network

and service layers and a lack of built-in facilities to support any nonbasic functionality. In order to move forward new architectures that can meet the research

and societal challenges and opportunities of Digital Society are needed. Incremental changes to existing architectures, which are enhancing the existing Internet,

are also of significant importance. Such new architectures, enhancements related artefacts would be based on: Emerging promising concepts,

which have the potential reach beyond current Internet core networking and servicing protocols, components, mechanisms and requirements.

Integration models enabling better incorporation and usage of the communicationcentric, information-centric, resource-centric, content-centric, service/computationcentric, context-centric faces and internet of things-centric facets

. 4 Part I: Future Internet Foundations: Architectural Issues Structures and infrastructures for control, configuration, integration, composition, organisation and federation.

Unification and higher degree of integration of the communication, storage, content and computation as the means of enabling change from capacity concerns towards increased and flexible capability with operation control.

Higher degree of virtualisation for all systems: applications, services, networks, storage, content, resources and smart objects.

Fusion of diverse design requirements, which include openness, economic viability, fairness, scalability, manageability, evolvability and programmability, autonomicity, mobility, ubiquitous access, usage,

The content of this area includes eight chapters covering some of the above architectural research in Future Internet.

The Towards a Future Internet Architecture chapter identifies the fundamental limitations of Internet, which are isolated not but strongly dependent on each other.

The transmission can be improved by utilising better data processing & handling and better data storage while the overall Internet performance would be improved significantly by control & self-*functions.

As an overall result this chapter proposes the following: extensions, enhancements and re-engineering of today's Internet protocols may solve several challenging limitations.

Yet, addressing the fundamental limitations of the Internet architecture is a multidimensional problem. Improvements in each dimension combined with a holistic approach of the problem space are needed.

The Towards In-Network Clouds in Future Internet chapter explores the architectural co-existence of new and legacy services and networks, via virtualisation of connectivity and computation resources and self management capabilities,

by fully integrating networking with cloud computing in order to create In-Network Clouds. It also presents the designs and experiments with a number of In-Network Clouds platforms

which have the aim to create a flexible environment for autonomic deployment and management of virtual networks and services as experimented with

Towards Scalable Future Internet Mobility chapter provides a comprehensive overview and review of the scalability problems of mobile Internet nowadays and to show how the concept of flat and ultra flat architectures emerges due to its suitability and applicability for the future Internet.

It also aims to introduce the basic ideas and the main paradigms behind the different flat networking approaches trying to cope with the continuously growing traffic demands.

The analysis of these areas guides the readers from the basics of flat mobile Internet architectures to the paradigm's complex feature set

and power creating a novel Internet architecture for future mobile communications. The Review and Designs of Federated Management in Future Internet Architectures chapter analyses issues about federated management targeting information sharing capabilities for heterogeneous infrastructure.

An inter-operable, extensible, Part I: Future Internet Foundations: Architectural Issues 5 reusable and manageable new Internet reference model is critical for Future Internet realisation and deployment.

The reference model must rely on the fact that high-level applications make use of diverse infrastructure representations and not use of resources directly.

So when resources are not being required to support or deploy services they can be used in other tasks or services.

and all these activities are developed under the umbrella of the federated management work in the future Internet.

The An Architectural Blueprint for a Real-world Internet chapter reviews a number of architectures developed in projects in the area of Real-world Internet (RWI), Internet of things (Iot),

and Internet Connected Objects. All of these systems are faced with very similar problems in their design with very limited interoperability among these systems.

The Towards a RESTFUL Architecture for Managing a Global Distributed Interlinked Data-Content-Information Space chapter analyses the concept of Content-Centric architecture, lying between the Web of Documents and the generalized Web of Data, in

which explicit data are embedded in structured documents enabling consistent support for the direct manipulation of information fragments.

uniform Web-based interface to distributed heterogeneous information management; it endows information fragments with collaboration-oriented properties, namely:

The A Cognitive Future Internet Architecture chapter proposes a novel Cognitive Framework as a reference architecture for the Future Internet (FI),

On one hand, it aims at achieving a full interoperation among the different entities constituting the ICT environment, by means of the introduction of Semantic Virtualization Enablers.

Future Internet Foundations: Architectural Issues The Title Model Ontology for Future Internet Networks chapter contributes to the use of ontologies in the future Internet, with the proposal of semantic formalization of the Entity Title Model.

It is suggested also the use of semantic representation languages in place of protocols. Alex Galis and Theodore Zahariadis J. Domingue et al.

) Future Internet Assembly, LNCS 6656, pp. 7 18,2011. The Author (s). This article is published with open access at Springerlink. com Towards a Future Internet Architecture Theodore Zahariadis1, Dimitri Papadimitriou2, Hannes Tschofenig3, Stephan Haller4, Petros Daras5

, George D. Stamoulis6, and Manfred Hauswirth7 1 Synelixis Solutions Ltd/TEI of Chalkida, Greece zahariad@{synelixis. com, teihal. gr} 2 Alcatel-lucent, Belgium dimitri. papadimitriou@alcatel-lucent

. com 3 Nokia Siemens Networks, Germany hannes. tschofenig@nsn. com 4 SAP, Germany stephan. haller@sap. com 5 Center of Research

and Technology Hellas/ITI, Greece daras@iti. gr 6. Athens University of Economics and Business,

In the near future, the high volume of content together with new emerging and mission critical applications is expected to stress the Internet to such a degree that it will possibly not be able to respond adequately to its new role.

and research initiatives worldwide to search for structural modifications to the Internet architecture in order to be able to face the new requirements.

This paper is based on the results of the Future Internet Architecture (FIARCH) group organized and coordinated by the European commission (EC)

and aims to capture the group's view on the Future Internet Architecture issue. Keywords:

Internet Architecture, Limitations, Processing, Handling, Storage, Transmission, Control, Design Objectives, EC FIARCH group. 1 Introduction The Internet has evolved from a remote access to mainframe computers and slow

Billions of people all over the world use the Internet for finding, accessing and exchanging information,

enjoying multimedia communications, taking advantage of advanced software services, buying and selling, keeping in touch with family and friends,

The success of the Internet has created even higher hopes and expectations for new applications and services

which the current Internet may not be able to support to a sufficient level. It is expected that the number 8 T. Zahariadis et al. of nodes (computers, terminals mobile devices, sensors, etc.

of the Internet will soon grow to more than 100 billion 1. Reliability, availability, and interoperability required by new networked services,

and this trend will escalate in the future. Therefore, the requirement of increased robustness, survivability, and collaborative properties is imposed to the Internet architecture.

In parallel, the advances in video capturing and content/media generation have led to very large amounts of multimedia content

, 3d videos, interactive environments, network gaming, virtual worlds, etc. compared to the quantity and type of data currently exchanged over the Internet.

Based on 2, out of the 42 Exabytes (1018) of consumer Internet traffic likely to be generated every month in 2014,56%will be due to Internet video,

while the average monthly consumer Internet traffic will be equivalent to 32 million people streaming Avatar in 3d, continuously, for the entire month.

All these applications create new demands and requirements, which to a certain extent can be addressed by means of over-dimensioning combined with the enhancement of certain Internet capabilities over time.

While this can be a satisfactory (although sometimes temporary) solution in some cases analyses have shown 3,

4 that increasing the bandwidth on the backbone network will not suffice due to new qualitative requirements concerning, for example, highly critical services such as ehealth applications, clouds of services and clouds of sensors, new social network

if the architecture and its properties might become the limiting factor of Internet growth and of the deployment of new applications.

the evolution of the Internet architecture is carried out by means of incremental and reactive additions 6, rather than by major and proactive modifications.

or richer functionality implying an architectural change define necessary but not sufficient conditions for such change in the Internet architecture and/or its components.

Indeed, the Internet architecture has shown since so far the capability to overcome such limits without requiring radical architectural transformation.

or designing a new Internet Architecture (if a new one is needed), it is necessary to demonstrate the fundamental limits of the current architecture 7. Thus,

scientists and researchers from both the industry and academia worldwide are working towards understanding these architectural limits so as to progressively determine the principles that will drive the Future Internet architecture that will adequately meet at least the abovementioned challenges EIFFEL,

The Future Internet as a global and common communication and distributed information system may be considered from various interrelated perspectives:

In Towards a Future Internet Architecture 9 Europe, a significant part of the Information and Communication Technology (ICT) of the Framework Program 7 is devoted to the Future Internet 14.

Though many proposals for a Future Internet Architecture have already been developed, no specific methodology to evaluate the efficiency

The purpose of this paper is to capture the view of the Future Internet Architecture (FIARCH) group organized and coordinated by the European commission.

and reached some understanding and agreement on the different types of limitations of the Internet and its architecture.

data structures, state machines) and the characterization of their interactions (i e. messages, calls, events, etc..We also qualify as a fundamental limitation of the Internet architecture a functional, structural,

or performance restriction or constraint that cannot be resolved effectively with current or clearly foreseen architectural paradigms as far as our understanding/knowledge goes.

so that this would in turn change the global properties of the Internet architecture (e g. separation of the locator and identifier role of IP ADDRESSES).

In the following, we use the term data to refer to any organized group of bits a k a. data packets, data traffic, information, content (audio, video, multimedia),

etc. and the term service to refer to any action performed on data or other services and the related Application programming interface (API).

2 Note however that this document does not take position on the localization and distribution of these APIS. 3 Analysis Approach Since its creation,

the Internet is driven by a small set of fundamental design principles rather than a formal architecture that is created on a whiteboard by a standardization or research group.

Moreover, the necessity for backwards compatibility and the trade-off between Internet redesign and proposing extensions,

enhancements and reengineering of today's Internet protocols are debated heavily. 1 Interested readers may also search for updated versions at the FIARCH site:

http://ec. europa. eu/information society/activities/foi/research/fiarch/index en. htm 2 The definition of service does not include the services offered by humans using the Internet 10 T. Zahariadis et al.

The emergence of new needs at both functional and performance levels, the cost and complexity of Internet growth, the existing

and performance limitations of the Internet's architectural principles and design model put the following elementary functionalities under pressure:

Processing/handling of data: refers to forwarders (e g. routers, switches, etc. computers (e g.,, terminals, servers, etc.

CPUS, etc. and handlers (software programs/routines) that generate and treat as well as query and access Data storage of data:

refers to memory, buffers, caches, disks, etc. and associated logical data structures. Transmission of data: refers to physical and logical transferring/exchange of data.

Control of processing, storage, transmission of systems and functions: refers to the action of observation (input), analysis,

and decision (output) whose execution affects the running conditions of these systems and functions. Note that by using these base functions,

the data communication function can be defined as the combination of processing, storage, transmission and control functions applied to data.

The term control is used here to refer to control functionality but also management functionality, e g. systems, networks, services, etc.

For each of the above functionalities, the FIARCH group has tried to identify and analyze the presumed problems and limitations of the Internet.

This work was carried out by identifying an extensive list of limitations and potentially problematic issues or missing functionalities

and then selecting the ones that comply with the aforementioned definition of a fundamental limitation. 3. 1 Processing

and Handling Limitations The fundamental limitations that have been identified in this category are: i. The Internet does not allow hosts to diagnose potential problems

and the network offers little feedback for hosts to perform root cause discovery and analysis. In today's Internet,

when a failure occurs it is often impossible for hosts to describe the failure (what happened?)

or selfish interests is detrimental to the cooperation between Internet users and providers. Non-intrusive and non-discriminatory means to detect misbehavior

while keeping open and broad accessibility to the Internet is a limitation that is crucial to overcome 16. ii.

Lack of data identity is damaging the utility of the communication system. As a result, data,

as an‘economic object',traverses the communication infrastructure multiple times, limiting its scaling, while lack of content‘property rights'(not only author-but also usage-rights) leads to the absence of a fair charging model. iii.

Towards a Future Internet Architecture 11 iv. Real-time processing. Though this is not directly related to the Internet Architecture itself,

the limited capability for processing data on a real-time basis poses limitations in terms of the applications that can be deployed over the Internet.

On the other hand, many application areas (e g. sensor networks) require real-time Internet processing at the edges nodes of the network. 3. 2 Storage Limitations The fundamental restrictions that have been identified in this category are:

i. Lack of context/content aware storage management: Data are associated not inherently with knowledge of their context.

This information may be available at the communication end-points (applications) but not when data are in transit.

So, it is not feasible to make efficient storage decisions that guarantee fast storage management, fast data mining and retrieval,

refreshing and removal optimized for different types of data 18. ii. Lack of inherited user and data privacy:

In case data protection/encryption methods are employed (even using asymmetric encryption and public key methods), data cannot be stored efficiently/handled.

On the other hand, lack of encryption, violates the user and data privacy. More investigations into the larger privacy and data protection ecosystem are required to overcome current limits of how current information systems deal with privacy and protection of information of users,

and develop ways to better respect the needs and expectations 30,31, 32 iii. Lack of data integrity, reliability and trust, targeting the security and protection of data;

this issue covers both unintended disclosure and damage to integrity from defects or failures, and vulnerabilities to malicious attacks. iv.

Lack of efficient caching & mirroring: There is no inherited method for on-path caching along the communication path

and mirroring of content compared to offpath caching that is currently widely used (involving e g. connection redirection).

Such methods could deal with issues like flash crowding, as the onset of the phenomenon will still cause thousands of cache servers to request the same documents from the original site of publication. 3. 3 Transmission Limitations The fundamental restrictions that have been identified in this category are:

i. Lack of efficient transmission of content-oriented traffic: Multimedia contentoriented traffic comprises much larger volumes of data as compared to any other information flow,

while its inefficient handling results in retransmission of the same data multiple times. Content Delivery Networks (CDN) and more generally architectures using distributed caching alleviate the problem under certain conditions

but can't extend to meet the Internet scale 19. Transmission from centralized locations creates unnecessary overheads

and can be far from optimal when massive amounts of data are exchanged. 12 T. Zahariadis et al. ii.

Lack of integration of devices with limited resources to the Internet as autonomous addressable entities.

Devices in environments such as sensor networks or even nano-networks/smart dust as well as in machine to machine-machine (M2m) environments operate with such limited processing,

storage and transmission capacity that only partly run the protocols necessary in order to be integrated in the Internet as autonomous addressable entities. iii.

Security requirements of the transmission links: Communications privacy does not only mean protecting/encrypting the exchanged data

but also not disclosing that communication took place. It is not sufficient to just protect/encrypt the data (including encryption of protocols/information/content,

tamper-proof applications etc) but also protect the communication itself, including the relation/interaction between (business

In the current Internet model, design of IP (and more generally communication) control components have so far being driven exclusively by

and (operational and system) cost of the Internet. Further, to maintain/sustain or even increase its value delivery over time,

the Internet will have to provide flexibility in its functional organization, adaptation, and distribution. Flexibility at run time is essential to cope with the increasing uncertainty (unattended and unexpected events) as well as breadth of expected events/running conditions for

Improper segmentation of data and control. The current Internet model segments (horizontally) data and control,

whereas from its inception the control functionality has a transversal component. Thus, on one hand, the IP functionality isn't limited anymore to the network layer,

Towards a Future Internet Architecture 13 share the same control instance. Hence, the hourglass model of the Internet does not account for this evolution of the control functionality

when considered as part of the design model. iii. Lack of reference architecture of the IP control plane.

The IP data plane is itself relatively simple but its associated control components are numerous and sometimes overlapping,

Addressing effectively the trade-off of network support without decreasing its scaling properties by requiring maintenance of per-flow state is one of the Internet's main challenges 16.3.5 Limitations That May Fall in More than One Category Certain fundamental limitations

of current Internet may fall in more than one category. Examples of such limitations include i. Traffic growth vs heterogeneity in capacity distribution:

Hosts connected to the Internet do not have the possibility to enforce the path followed by their traffic.

On the other hand, as the Internet enables any-to-any connectivity, there is no effective means to predict the spatial distribution of the traffic within a timescale that would allow providers to install needed capacity

The huge number of (mobile) terminals combined with a sudden peak in demand for a particular piece of data may result in phenomena that cannot be handled;

The amount of foreseen data and information5 requires significant processing power/storage/bandwidth for indexing/crawling

and (distributed) querying and also solutions for large scale/real-time data mining/social network analysis, so as to achieve successful retrieval and integration of information from an extremely high numer of sources across the network.

All the aforementioned issues imply the need for addressing new architectural challenges capable to cope with the fast and scalable identification and discovery of and access to data.

v. Security of the whole Internet Architecture. The Internet architecture is not intrinsically secure and is based on add-ons to, e g. protocols,

to secure itself. The consequence is that protocols may be secure but the overall architecture is not selfprotected against malicious attacks. vi.

Support of mobility when using IP ADDRESS as both network and host identifier but also TCP connection identifier results in Transmission control protocol (TCP) connection continuity problem.

(when the wireless link cannot be conditioned to properly control its error rate or due to transient wireless link interruption in areas of poor coverage), rendering the typical reaction of congestion control mechanism of TCP inappropriate.

and affecting several base functions again. 4 Design Objectives The purpose of this section is to document the design objectives that should be met by the Internet architecture.

but also technological expectations to be met by the Internet as global and common information communication system. High-level objectives are documented in 15.

By low-level design objectives, we mean here the functional and performance properties as well as the structural and quality properties that the architecture of this global and 5 Eric Schmidt, the CEO of Google,

the world's largest index of the Internet, estimated the size at around 5 million terabytes of data (2005).

Eric commented that Google has indexed roughly 200 terabytes of that is 0, 004%of the total size.

Towards a Future Internet Architecture 15 common information communication system is expected to meet. From the previous sections, some of low-level objectives are met

and others are not by the (present) architecture of the Internet. We also emphasize here that these objectives are shared commonly by the Internet community at large The remaining part of this Section translates a first analysis of the properties that should be met by the Internet architecture starting from the initial of objectives as enumerated in various references (see 27

28,29. One of the key challenges is thus to determine the necessary addition/improvement of current architecture principles

the Internet architecture has been structured around eight foundational objectives: i) to connect existing networks, ii) survivability,

underlines that the Internet architecture needs to be able to scale to 109 IP networks recognizing the need to add scalability as a design objective.

In this context, the followed approach consists of starting from the existing Internet design objectives compared to the approach that would consist of applying a tabula rasa approach

, completely redefine from scratch the entire set of Internet design objectives. Based on previous sections, the present section describes the design objectives that are met currently,

Accessibility (open and by means of various/heterogeneous wireless/radio and wired interfaces) to the communication network but also to heterogeneous data, applications,

Accessibility and nomadicity are addressed currently by current Internet architecture. On the other hand, mobility is realized still in most cases by means of dedicated/separated architectural components instead of Mobile IP. see Subsection 3. 5. Point 6 Accountability of resource usage and security without impeding

in the current Internet this service is the connectivity even if the notion of service is embedded not in the architectural model of the Internet:

initially addressed but loosing ground. Distribution of processing, storage, and control functionality and autonomy (organic deployment):

concerning storage and processing, several architectural enhancements might be required, e g. for the integration of distributed but heterogeneous data and processes. 16 T. Zahariadis et al.

referring here to the capacity of the Internet to perform in accordance to what it is expected to deliver to the end-user/hosts while coping with a growing number of users with increasing heterogeneity in applicative communication needs.

and associated data traffic such as non/real-time streams, messages, etc.,independently of the shared infrastructure partitioning/divisions,

addressed and to be reinforced (migration of mobile network to IPV6 Internet, IPTV moving to Internet TV, etc.

but simplicity seems to be progressively decreasing see Section 3. 4 Point 3. Note that simplicity is added explicitly as a design objective to-at least-prevent further deterioration of the complexity of current architecture (following the Occam's razor principle).

and Subsection 3. 5, Point 4. 5 Conclusions In this article we have identified fundamental limitations of Internet architecture following a systematic investigation thereof from a variety of different viewpoints.

The transmission can be improved by utilizing better data processing and handling (e g. network coding, data compression, intelligent routing) and better data storage (e g. network/terminals caches,

data centers/mirrors etc. while the overall Internet performance would be improved significantly by control and self-*functions.

As an overall finding we may conclude the following: Extensions, enhancements and re-engineering of today's Internet protocols may solve several challenging limitations.

Yet, addressing the fundamental limitations of the Internet architecture is a multidimensional and challenging research topic.

While improvements are needed in each dimension, these should be combined by undertaking a holistic approach of the problem space.

Acknowledgements. This article is the based on the work that has been carried out by the EC Future Internet Architecture (FIARCH) group (to

which the authors belong), which is coordinated by the EC FP7 Coordination and Support Actions (CSA) projects Towards a Future Internet Architecture 17 in the area of Future Internet:

Nextmedia, IOT-I, SOFI, EFFECTS+,EIFFEL, Chorus+,SESERV and Paradiso 2, and supported by the EC Units D1:

Future Networks, D2: Networked Media Systems, D3: Software & Service Architectures & Infrastructures, D4: Networked Enterprise & Radio frequency identification (RFID) and F5:

Trust and Security. The authors would like to acknowledge and thank all members of the group for their significant input and the EC Scientific Officers Isidro Laso Ballesteros, Jacques Babot, Paulo De Sousa, Peter Friess, Mario Scillia

, Arian Zwegers for coordinating the activities. The authors would like also to acknowledge the FI architectural work performed under the project FP7 COAST ICT-248036 COAST.

The Internet and its architecture have grown in evolutionary fashion from modest beginnings, rather than from a Grand Plan 7 Li, T. ed.:

Design Goals for Scalable Internet Routing. Work in progress, draft-irtf-rrgdesign-goals-02.sep 2010) 8 http://www. nsf. gov/pubs/2010/nsf10528/nsf10528. htm 9 http://www. nsf

. pdf 14 http://www. future-internet. eu/15 FIARCH Group: Fundamental Limitations of Current Internet and the path to Future Internet (December 2010) 16 Perry, D.,Wolf, A.:

Foundations for the Study of Software Architecture. ACM SIGSOFT Software engineering Notes 17,4 (1992) 17 Papadimitriou, D.,et al.

eds.):) Open Research Issues in Internet Congestion Control. Internet Research Task force (IRTF), RFC 6077 (February 2011) 18 Akhlaghi, S.,Kiani, A.,Reza Ghanavati, M.:

Cost-bandwidth tradeoff in distributed storage systems (published on-line. ACM Computer Communications 33 (17), 2105 2115 (2010) 19 Freedman, M.:

Experiences with Coralcdn: A Five-Year Operational View. In: Proc. 7th USENIX/ACM Symposium on Networked Systems Design and Implementation (NSDI'10) San jose, CA (May 2010) 18 T. Zahariadis et al. 20 Dobson, S.,et al.:

A survey of autonomic communications. ACM Transactions on Autonomous and Adaptive Systems (TAAS) 1 (2), 223 259 (2006) 21 Gelenbe, E.:

Steps toward self-aware networks. ACM Communications 52 (7), 66 75 (2009) 22 Evolving the Internet, Presentation to the OECD (March 2006), http://www. cs. ucl. ac. uk/staff

/m. handley/slides/23 Meyer, D.,et al.:Report from the IAB Workshop on Routing and Addressing, IETF, RFC 4984 (Sep. 2007) 24 Mahonen, P. ed.),Trossen, D.,Papadimitrou, D.,Polyzos, G.,Kennedy, D.:

Invigorating the Future Internet Debate. ACM SIGCOMM Computer Communication Review 39 (5)( 2009) 26 Eggert, L.:

Quality-of-Service: An End System Perspective. In: MIT Communications Futures Program Workshop on Internet Congestion Management, Qos,

and Interconnection, Cambridge, MA, USA, October 21-22 (2008) 27 Ratnasamy, S.,Shenker, S.,Mccanne, S.:

Towards an evolvable internet architecture. SIGCOMM Comput. Commun. Rev. 35 (4), 313 324 (2005) 28 Cross-ETP Vision Document, http://www. future-internet. eu/fileadmin/documents/reports/Cross-ETPS FI VISION DOCUMENT V1 0

. pdf 29 Clark, D d.:The Design Philosophy of the DARPA Internet Protocols, Proc SIGCOMM 88 (reprinted in ACM CCR 25 (1), 102-111,1995.

ACM CCR 18 (4), 106 114 (1988) 30 Saltzer, J. H.,Reed, D. P.,Clark, D d.:

Architectural Principles of the Internet, Internet Engineering Task force (IETF), RFC 1958 (July 1996) 32 Krishnamurthy, B.:

ACM SIGCOMM Computer Communication Review (Oct. 2010), http://www2. research. att. com/bala/papers/ccr10-priv. pdf 33 W3c Workshop

on Privacy for Advanced Web APIS 12/13 July 2010, London (2010), http://www. w3. org/2010/api-privacy-ws/report. html 34

Workshop on Internet Privacy, co-organized by the IAB, W3c, MIT, and ISOC, 8 and 9 december (2010), http://www. iab. org/about/workshops/privacy/35 Clark, D.,et al.:

Towards the Future Internet Architecture, Internet Engineering Task force (IETF; RFC 1287 (December 1991) 36 http://www. iso. org/iso/iso technical committee. html?

://www. fp7-eiffel. eu/EULER http://www. euler-project. eu/Iot-A http://www. iot-a. eu/nextmedia http://www. fi-nextmedia

) Future Internet Assembly, LNCS 6656, pp. 19 33,2011. The Author (s). This article is published with open access at Springerlink. com. Towards In-Network Clouds in Future Internet Alex Galis1, Stuart Clayman1, Laurent Lefevre2, Andreas Fischer3, Hermann

de Meer3, Javier Rubio-Loyola4, Joan Serrat5, and Steven Davy6 1 University college London, United kingdom, {a. galis, s. clayman}@ ee. ucl. ac. uk 2 INRIA, France, laurent. lefevre@ens-lyon. fr 3 University of Passau, Germany {andreas

One of the key aspect fundamentally missing from the current Internet infrastructure is advanced an service networking platform and facilities,

which take advantage of flexible sharing of available connectivity, computation, and storage resources. This paper aims to explore the architectural co-existence of new and legacy services and networks, via virtualisation of connectivity and computation resources and self management capabilities,

by fully integrating networking with cloud computing in order to create In-Network Clouds. It also presents the designs and experiments with a number of In-Network Clouds platforms

which have the aim to create a flexible environment for autonomic deployment and management of virtual networks and services as experimented with

In-Network Clouds, Virtualisation of Resources, Self management, Service plane, Orchestration plane and Knowledge plane. 1 Introduction The current Internet has been founded on a basic architectural premise, that is:

The simplicity of the current Internet has pushed complexity into the endpoints, and has allowed impressive scale in terms of interconnected devices.

Internet applications increasingly require a combination of capabilities from traditionally separate technology domains to deliver the flexibility

Internet use is expected to grow massively over the next few years with an order of magnitude more Internet services

the interconnection of smart objects from the Internet of things, and the integration of increasingly demanding enterprise and societal applications.

The Future Internet research and development trends are covering the main focus of the current Internet, which is connectivity,

As such, the Future Internet covers the complete management and full lifecycle of applications, services, networks and infrastructures that are constructed primarily by recombining existing elements in new and creative ways.

The aspects which are fundamentally missing from the current Internet infrastructure include the advanced service networking platforms and facilities,

which take advantage of flexible sharing of available resources (e g. connectivity, computation, and storage resources). This paper aims to explore the architectural co-existence of new and legacy services and networks, via virtualisation of resources and self management capabilities,

by fully integrating networking 4, 8, 10,15 with cloud computing 6, 7, 9 in order to produce In-Network Clouds.

It also presents the designs and experiments with a number of In-Network Clouds platforms 9, 10,

modifications to the existing Internet are limited now to simple incremental updates and deployment of new technology is next to impossible and very costly.

In-Network virtualisation provides flexibility, promotes diversity, and promises security and increased manageability. We define In-Network clouds as an integral part of the differentiated Future Internet architecture,

which supports multiple computing clouds from different service providers operating on coexisting heterogeneous virtual networks

and sharing a common physical substrate of communication nodes and servers managed by multiple infrastructure providers.

By decoupling service providers from infrastructure providers and by integrating computing clouds with virtual networks the In-Network clouds introduce flexibility for change.

In-Network Network and Service Clouds can be represented by a number of distributed management systems described with the help of five abstractions:

Virtualisation Plane (VP), Management Plane (MP), Knowledge Plane (KP), Service Plane (SP), and Orchestration Plane (OP) as depicted in Fig. 1. These planes are new higher-level artefacts,

used to make the Future Internet of Services more intelligent, with embedded management functionality. At a logical level, the VMKSO planes gather observations, constraints and assertions,

and servers Towards In-Network Clouds in Future Internet 21 Fig. 1. In-Network Cloud Resources within the network.

Together these distributed systems form a software-driven network control infrastructure that will run on top of all current networks (i e. fixed

wireless, and mobile networks) and service physical infrastructures in order to provide an autonomic virtual resource overlay. 2. 1 Service Plane Overview The Service Plane (SP) consists of functions for the automatic (redeployment of new

It supervises and it integrates all other planes'behaviour ensuring integrity of the Future Internet management operations.

These components could have direct interworking with control algorithms situated in the control plane of the Internet

(i e. to provide real time reaction), and interworking with other management functions (i e. to provide near real time reaction).

Since each domain may have different SLAS, security and Towards In-Network Clouds in Future Internet 23 administrative policies,

The governance functionality of the OP monitors the consistency of the AMSS'actions, it enforces the high level policies

This implies that the Orchestration Plane may use very local knowledge to deserve a real time control as well as a more global knowledge to manage some long-term processes like planning. 2. 3 Virtualisation Plane Overview Virtualisation hides the physical characteristics 14,16 of the computing

This paper uses system virtualisation to provide virtual services and resources. System virtualisation separates an operating system from its underlying hardware resources;

resource virtualisation abstracts physical resources into manageable units of functionality. For example, a single physical resource can appear as multiple virtual resources (e g.,

, the concept of a virtual router, where a single physical router can support multiple independent routing processes by assigning different internal resources to each routing process;

alternatively, multiple physical resources can appear as a single physical resource (e g.,, when multiple switches are stacked so that the number of switch ports increases,

Virtualisation enables optimisation of resource utilisation. However, this optimisation is confined to inflexible configurations within a single administrative domain.

This paper extends contemporary virtualisation approaches and aims at building an infrastructure in which virtual machines can be relocated dynamically to any physical node or server regardless of location, network,

and storage configurations and of administrative domain. The virtualisation plane consists of software mechanisms to abstract physical resources into appropriate sets of virtual resources that can be organised by the Orchestration Plane to form components (e g.,

, increased storage or memory devices (e g.,, a switch with more ports or even networks. The organisation is done

in order to realise a certain business goal or service requirement. Two dedicated interfaces are needed: the vspi and the vcpi (Virtualisation System Programming interface and Virtualisation Component Programming interface, respectively.

A set of control loops is formed using the vspi and the vcpi, as shown in Figure 2. 24 A. Galis et al.

Fig. 2. Virtualisation Control Loop Virtualisation System Programmability Interface (vspi. The vspi is used to enable the Orchestration Plane

Virtualisation Component Programming interface (vcpi. Each physical resource has associated an and distinct vcpi. The vcpi is fulfilling two main functions:

and to request virtual resources to be constructed from that physical resource by the vcpi of the Virtualisation Plane.

The vcpi also provides monitoring information from the virtual resources back to the AMS that Towards In-Network Clouds in Future Internet 25 controls that physical resource.

Note that the AMS is responsible for obtaining management data describing the physical resource. The vcpi is responsible for providing dynamic management data to its governing AMS that states how many virtual resources are currently instantiated,

and how many additional virtual resources of what type can be supported. 2. 4 Knowledge Plane Overview The Knowledge Plane was proposed by Clark et al. 1 as a new dimension to a network architecture, contrasting with the data and control planes;

its purpose is to provide knowledge and expertise to enable the network to be self-monitoring, selfanalysing,

A narrow functionality Knowledge Plane (KP), consisting of context data structured in information models and ontologies,

The KP brings together widely distributed data collection, wide availability of that data, and sophisticated and adaptive processing or KP functions, within a unifying structure.

Knowledge extracted from information/data models forms facts. Knowledge extracted from ontologies is used to augment the facts,

which is used then to transform received data into a common form that enables it to be managed.

Furthermore, context-aware networking enables new types of applications and services in the future Internet. Context Information Services.

i) the Context Executive (CE) Module which interfaces with other entities/context clients,(ii) the Context Processing (CP) Module which implements the core internal operations related to the context processing

In general, the CE module is responsible for the communication of the CISP with the other management Towards In-Network Clouds in Future Internet 27 applications/components and the CP module for the optimisation of the context information.

The Context Information Base (CIB) provides flexible storage capabilities, in support of the Context Executive and Context Processor modules.

they monitor hardware and software for their state, present their capabilities, or collect configuration parameters.

A monitoring mechanism and framework was developed to gather measurements from relevant physical and virtual resources and CCPS for use within the CISP.

, the number of CPUS,(ii) N-time queries, which collect information periodically, and (iii) continuous queries that monitor information in an ongoing manner.

CCPS should be located near the corresponding sources of information in 28 A. Galis et al. order to reduce management overhead.

Filtering rules based on accuracy objectives should be applied at the CCPS, especially for the N-time and continuous queries, for the same reason.

This can include common operations such as getting the state of a server with its CPU

and number of bytes coming in and out, or getting the state of disks on a system presenting the total volume, free space,

In our implementation, each sensor runs in its own thread allowing each one to collect data at different rates

We note that the monitoring information retrieval is handled by the Virtualisation Plane. The reader collects the raw measurement data from all of the sensors of a CCP.

The collection can be done at a regular interval or as an event from the sensor itself.

The reader collects data from many sensors and converts the raw data into a common measurement object used in the CISP Monitoring framework.

The format contains meta-data about the sensor and the time of day, and it contains the retrieved data from the sensor.

The filter takes measurements from the reader and can filter them out before they are sent on to the forwarder.

In our case, the filtering percentage matches the accuracy objective of the management application requesting the information.

and to set the rate at which they collect data;(ii) the filtering process, by changing the filter

iii) the forwarder, by changing the attributes of the network (such as IP ADDRESS and port) that the ICP is connected to.

which can measure attributes from CPU, memory, and network components of a server host, were created.

We can also measure the same attributes of virtualised hosts by interacting with a hypervisor to collect these values.

Towards In-Network Clouds in Future Internet 29 2. 5 Management Plane Overview The Management Plane is a basic building block of the infrastructure,

is responsible for the optimal placement and continuous migration of virtual routers into hosts (i e.,, physical nodes and servers) subject to constraints determined by the Orchestration Plane.

The Management Plane is designed to meet the following functionality: Embedded (Inside) Network functions: The majority of management functionality should be embedded in the network

which run on top of all current networks (i e. fixed, wireless and mobile networks) and service physical infrastructures.

It monitors the network and operational context as well as internal operational network state in order to assess if the network current behaviour serve its service purposes.

and continuous migration of virtual routers into hosts (i e. physical nodes and servers) subject to constraints determined by the Orchestration Plane.

Mapping logic enables the data stored in models to be transformed into knowledge and combined with knowledge stored in ontologies to provide a context-sensitive assessment of the operation of one or more virtual resources.

and issued as open source 10, which aims to create a highly open and flexible environment for In-Network Clouds in Future Internet.

They are described briefly herewith. Full design and implementation of all software platforms are presented in 10. vcpi (Virtual Component Programming interface is the VP's main component dealing with the heterogeneity of virtual resources

and enabling programmability of network elements In each physical node there is an embedded vcpi, which is aware of the structure of the virtual resources,

Towards In-Network Clouds in Future Internet 31 CISP (Context Information Service Platform) is the KP's main component supported by a distributed monitoring platform for resources & components.

also part of the KP, provides functionality to add powerful and flexible monitoring facilities to system clouds (virtualisation of networks and services.

The framework provides data sources, data consumers, and a control strategy. In a large distributed system there may be hundreds or thousands of measurement probes,

which can generate data. APE (Autonomic Policy-based Engine), a component of the MP, supports contextaware policy-driven decisions for management and orchestration activities.

control and management of programmable or active sessions over virtual entities, such as servers and routers.

RNM (Reasoning and Negotiation Module), a core element of the KP, which mediates and negotiates between separate federated domains.

and validated on 2 testbeds enabling experimentation with thousands of virtual machines: V3 UCL's Experimental Testbed located in London consisting of 80 cores with a dedicated 10 Gbits/s infrastructure

and Grid5000-an Experimental testbed located in France consisting of 5000 cores and linked by a dedicated 10 Gbits/s infrastructure.

Validation and performance analysis are described fully in 13. Demonstrations are available at: http://clayfour. ee. ucl. ac. uk/demos/and they are used for:

Autonomic service provisioning on In-Network Clouds (Service Computing Clouds. 4 Conclusion This work has presented the design of an open software networked infrastructure (In-Network Cloud) that enables the composition of fast and guaranteed services in an efficient manner,

and the execution of these services in an adaptive way taking into 32 A. Galis et al. account better shared network

and service resources provided by an virtualisation environment. We have described also the management architectural and system model for our Future Internet,

which were described with the help of five abstractions and distributed systems the OSKMV planes: Virtualisation Plane (VP), Management Plane (MP), Knowledge Plane (KP), Service Plane (SP) and Orchestration Plane (OP). The resulting software-driven control network

infrastructure was exercised fully and relevant analysis on network virtualisation and service deployments were carried out on a large-scale testbed.

Virtualising physical network and server resources has served two purposes: Managing the heterogeneity through introduction of homogeneous virtual resources and enabling programmability of the network elements.

The flexibility gained through this approach helps to adapt the network dynamically to both unforeseen and predictable changes in the network.

A vital component of such a virtualisation approach is a common management and monitoring interface of virtualised resources.

Such an interface has exported management and monitoring functions that allow management components to control the virtual resources in a very fine-grained way through a single, well defined interface.

this interface can then form the basis for new types of applications and services in the future Internet.

This work was undertaken partially in the context of the FP7-EU Autonomic Internet 10 and the RESERVOIR 9 research projects,

Future Generation Internet Architecture, http://www. isi. edu/newarch/2. Galis, A.,et al.:Management and Service-aware Networking Architectures (MANA) for Future Internet Position Paper:

System Functions, Capabilities and Requirements. Invited paper IEEE Chinacom09 26-28, Xi'an, China (August 2009), http://www. chinacom. org/2009/index. html 3. Rubio-Loyola

Platforms and Software systems for an Autonomic Internet. IEEE Globecom 2010; 6-10 dec.,, Miami, USA (2010) 4. Galis, A.,et al.:

Management Architecture and Systems for Future Internet Networks. In: Towards the Future Internet, IOS Press, Amsterdam (2009) 5. Chapman, C.,et al.:

Software Architecture Definition for On-demand Cloud Provisioning. ACM HPDC, 21-25, Chicago hpdc2010. eecs. northwestern. edu (June 2010) 6. Rochwerger, B.,et al.:

An Architecture for Federated Cloud computing. In: Cloud computing, Wiley, Chichester (2010) 7. Chapman, C.,et al.:Elastic Service Management in Computational Clouds. 12th IEEE/IFIP NOMS2010/Cloudman 2010 19-23 april, Osaka (2010) http://cloudman2010. lncc. br/Towards In-Network

Clouds in Future Internet 33 8. Clayman, S.,et al.:Monitoring Virtual Networks with Lattice. NOMS2010/Manfi 2010-Management of Future Internet 2010;

19-23 april, Osaka, Japan (2010), http://www. manfi. org/2010/9. RESERVOIR project, http://www. reservoir-fp7. eu 10.

Autoi project http://ist-autoi. eu 11. Clark, D.,Partridge, C.,Ramming, J. C.:and, J. T. Wroclawski A Knowledge Plane for the internet.

In: Proceedings of the 2003 Conference on Applications, Technologies, Architectures, and Protocols For Computer Communications (Karlsruhe, Germany, SIGCOMM'03, Karlsruhe, Germany, August 25 29,2003, pp. 3 10.

ACM, New york (2003) 12. Jennings, B.,Van der Meer, S.,Balasubramaniam, S.,Botvich, D.,Foghlu, M.,Donnelly, W.,Strassner, J.:

Towards autonomic management of communications networks. IEEE Communications Magazine 45 (10), 112 121 (2007) 13. Deliverable D6. 3 Final Results Autoi Approach http://ist-autoi. eu/14.

A Survey of Network Virtualization. Journal Computer networks: The International Journal of Computer and Telecommunications Networking 54 (5)( 2010) 15.

Galis, A.,Denazis, S.,Bassi, A.,Berl, A.,Fischer, A.,de Meer, H.,Strassner, J.,Davy, S.,Macedo, D.,Pujolle, G.,Loyola, J. R

.,Serrat, J.,Lefevre, L.,Cheniour, A.:Management Architecture and Systems for Future Internet Networks. In:

Towards the Future Internet A European Research Perspective, p. 350. IOS Press, Amsterdam (2009), http://www. iospress. nl/16.

Berl, A.,Fischer, A.,De Meer, H.:Using System Virtualization to Create Virtualized Networks. Electronic communications of the EASST 17,1 12 (2009), http://journal. ub. tu-berl. asst/article/view/218/219 J. Domingue et al.

Eds.):) Future Internet Assembly, LNCS 6656, pp. 35 50,2011. The Author (s). This article is published with open access at Springerlink. com. Flat Architectures:

Towards Scalable Future Internet Mobility László Bokor, Zoltán Faigl, and Sándor Imre Budapest University of Technology and Economics department of Telecommunications Mobile Communication and Computing Laboratory Mobile Innovation Centre Magyar Tudosok krt. 2, H-1117

, Budapest Hungary {goodzi, szlaj, imre}@ mcl. hu Abstract. This chapter is committed to give a comprehensive overview of the scalability problems of mobile Internet nowadays

and to show how the concept of flat and ultra flat architectures emerges due to its suitability and applicability for the future Internet.

It also aims to introduce the basic ideas and the main paradigms behind the different flat networking approaches trying to cope with the continuously growing traffic demands.

The discussion of the above areas will guide the readers from the basics of flat mobile Internet architectures to the paradigm's complex feature set

and power creating a novel Internet architecture for future mobile communications. Keywords: mobile traffic evolution, network scalability, flat architectures, mobile Internet, IP mobility, distributed and dynamic mobility management 1 Introduction Mobile Internet has started recently to become a reality

for both users and operators thanks to the success of novel, extremely practical smartphones, portable computers with easy-to-use 3g USB modems and attractive business models.

Based on the current trends in telecommunications, vendors prognosticate that mobile networks will suffer an immense traffic explosion in the packet switched domain up to year 2020 1 4

. In order to accommodate the future Internet to the anticipated traffic demands, technologies applied in the radio access

and core networks must become scalable to advanced future use cases. There are many existing solutions aiming to handle the capacity problems of current mobile Internet architectures caused by the mobile traffic data evolution.

Reserving additional spectrum resources is the most straightforward approach for increasing the throughput of the radio access,

and also spectrum efficiency can be enhanced thanks to new wireless techniques (e g.,, High Speed Packet Access, and Long term Evolution.

Heterogeneous systems providing densification and offload of the macrocellular network throughout pico, femtocells and relays or Wifi/Wimax interfaces also extend the radio range.

However, the deployment of novel technologies providing higher radio throughput (i e.,, higher possible traffic rates) easily generates new 36 L. Bokor, Z. Faigl,

and S. Imre usages and the traffic increase may still accelerate. Since today's mobile Internet architectures have been designed originally for voice services

and later extended to support packet switched services only in a very centralized manner, the management of this ever growing traffic demand is quite hard task to deal with.

Scalability of traffic, network and mobility management functions has become one of the most important questions of the future Internet.

However, the strongly centralized nature of current and planned mobile Internet standards (e g.,, the ones maintained by the IETF

or by the collaboration of 3gpp) prevents cost effective system scaling for the novel traffic demands.

flat and fully distributed mobile architectures are gaining more and more attention today. The goal of this chapter is to provide a detailed introduction to the nowadays emerging scalability problems of the mobile Internet

and also to present a state of the art overview of the evolution of flat and ultra flat mobile communication systems.

In order to achieve this we first introduce the issues relating to the continuously growing traffic load inside the networks of mobile Internet providers in Section 2. Then,

in Section 3 we present the main evolutionary steps of flat architectures by bringing forward the most important schemes,

and Scalability Problems of the Mobile Internet 2. 1 Traffic Evolution Characteristics of the Mobile Internet One of the most important reasons of the traffic volume increase in mobile telecommunications is demographical.

over 25%of the global population this means about two billion people are using the Internet.

the number of wireless broadband subscriptions is exceed about to the total amount of fixed broadband subscriptions and this development Flat Architectures:

Towards Scalable Future Internet Mobility 37 becomes even more significant considering that the volume of fixed broadband subscriptions is gathering much slower.

The expansion of wireless broadband subscribers not only inflates the volume of mobile traffic directly, but also facilitates the growth in broadband wireless enabled terminals.

However, more and more devices enable mobile access to the Internet, only a limited part of users is attracted

or open to pay for the Wireless internet services meaning that voice communication will remain the dominant mobile application also in the future.

Despite this and the assumption of 5 implying that the increase in the number of people potentially using mobile Internet services will likely saturate after 2015 in industrialized countries

the mobile Internet subscription growth potential will be kept high globally by two main factors. On one hand the growth of subscribers continues unbrokenly in the developing markets:

mobile broadband access through basic handhelds will be the only access to the Internet for many people in Asia/Pacific.

On the other hand access device, application and service evolution is expected also to sustain the capability of subscriber growth.

it is foreseen that due to the development of data-hungry entertainment services like television/radio broadcasting and Vod,

66%of mobile traffic will be video by 2014 2. A significant amount of this data volume will be produced by mobile Web-browsing

, Youtube. Cisco also forecasts that the total volume of video (including IPTV, Vod, P2p streaming, interactive video, etc.

and mobile) by the year 2012, producing a substantial increase of the overall mobile traffic of more than 200%each year 7. Video traffic is anticipated also to grow so drastically in the forthcoming years that it could overstep Peer-to-peer (P2p) traffic 4. Emerging web technologies (such as HTML5

), the increasing video quality requirements (HDTV, 3d, SHV) and special application areas (virtual reality experience sharing

and gaming) will further boost this process and set new challenges to mobile networks. Since video and related entertainment services seems to become dominant in terms of bandwidth usage

special optimization mechanisms focusing on content delivery will also appear in the near future. The supposed evolution of Content Delivery Networking (CDN)

and smart data caching technologies might have further impact on the traffic characteristics and obviously on mobile architectures.

Another important segment of mobile application and service evolution is social networking. As devices, networks and modes of communications evolve,

In the future, social networking might evolve even further, like to cover broader areas of personal communication in a more integrated way,

or to put online gaming on the next level deeply impregnated with social networking and virtual reality. Even though video seems to be a major force behind the current traffic growth of the mobile Internet,

there is another emerging form of communications called M2m (Machine to machine-Machine) which has the potential to become the leading traffic contributor in the future.

M2m sessions accommodate end-to-end communicating devices 38 L. Bokor, Z. Faigl, and S. Imre without human intervention for remote controlling,

monitoring and measuring, road safety, security/identity checking, video surveillance, etc. Predictions state that there will be 225 million cellular M2m devices by 2014 with little traffic per node but resulting significant growth in total,

mostly in uplink direction 3. The huge number of sessions with tiny packets creates a big challenge for the operators.

As a summary we can state that the inevitable mobile traffic evolution is foreseen thanks to the following main factors:

growth of the mobile subscriptions, evolution of mobile networks, devices, applications and services, and significant device increase potential resulted by the tremendous number of novel subscriptions for Machine to machine-Machine communications. 2. 2 Scalability Problems of the Mobile Internet Existing wireless telecommunication

infrastructures are prepared not to handle this traffic increase, current mobile Internet was designed not with such requirements in mind:

mobile architectures under standardization (e g.,, 3gpp, 3gpp2, Wimax Forum) follow a centralized approach which cannot scale well to the changing traffic conditions.

On one hand user plane scalability issues are foreseen for anchor-based mobile Internet architectures, where mechanisms of IP ADDRESS allocation and tunnel establishment for end devices are managed by high level network elements,

called anchor points (GGSN in 3gpp UMTS, PDN GW in SAE, and CSN for Wimax networks).

Each anchor point maintains special units of information called contexts, containing binding identity, tunnel identifier, required Qos,

etc. on a per mobile node basis. These contexts are updated continuously and used to filter

and route user traffic by the anchor point (s) towards the end terminals and vice versa. However, network elements (hence anchor points too) are limited in terms of simultaneous active contexts.

and access layer provides easy service convergence in current mobile Internet architectures but introduces additional complexity regarding session establishment procedures

, Policy and Charging Control architecture by 3gpp) to achieve interaction between the two levels during session establishment, modification and release routines.

Towards Scalable Future Internet Mobility 39 As a consequence, architectural changes are required for dealing with the ongoing traffic evolution:

future mobile networks must specify architecture optimized to maximize the end-user experience, minimize CAPEX/OPEX, energy efficiency, network performance,

and to ensure mobile networks sustainability. 3 Evolution of Flat Architectures 3. 1 Evolution of the Architecture of 3gpp Mobile networks Fixed networks were firstly subject to similar scalability problems.

The evolution of DSL access architecture has shown in the past that pushing IP routing and other functions from the core to the edge of the network results in sustainable network infrastructure.

The same evolution was started to happen within the wireless telecommunication and mobile Internet era. The 3gpp network architecture specifications having the numbers 03.02 8 and 23.002 9 show the evolution of the 3gpp network from GSM Phase 1 published in 1995 until the Evolved Packet System (EPS

) specified in Release 8 in 2010. The core part of EPS called Evolved Packet Core (EPC) is extended continuously with new features in Release 10 and 11.

The main steps of the architecture evolution are summarized in the followings. Fig. 1 illustrates the evolution steps of the packet-switched domain,

In Phase 1 (1995) the basic elements of the GSM architecture have been defined. The reasons behind the hierarchization and centralization of the GSM architecture were both technical and economical.

Primarily it offloaded the switching equipments (crossbar switch or MSC. In parallel, existing ISDN switches could be reused as MSCS

Fig. 1. The evolution of the packet-switched domain of the 3gpp architecture, including the main user plane anchors in the RAN and the CN. 40 L. Bokor, Z. Faigl,

Release 1999 (2002) describes the well known UMTS architecture clearly separating the CS and PS domains.

Seeing that UMTS was designed to be the successor of GSM, it is not strange that the central anchors remained in place in 3g and beyond.

Progress of mobile and wireless communication systems introduced some fundamental changes. The most drastic among them is that IP has become the unique access protocol for data networks

and the continuously increasing future wireless traffic is also based on packet data (i e.,Internet communication.

Due to the collateral effects of this change a convergence procedure started to introduce IP-based transport technology in the core and backhaul network:

Release 4 (2003) specified the Media gateway function, Release 5 (2003) introduced the IP Multimedia Subsystem (IMS) core network functions for provision of IP services over the PS domain,

while Release 6 standardized WLAN interworking and Multimedia Broadcast Multicast Service (MBMS). With the increasing IP-based data traffic flattening hierarchical and centralized functions became the main driving force in the evolution of 3gpp network architectures.

Release 7 (also called Internet HSPA, 2008) supports the integration of the RNC with the Nodeb providing a one node based radio access network.

Another architectural enhancement of this release is the elaboration of Direct Tunnel service 10 11.

Direct Tunnel allows to offload user traffic from SGSN by bypassing it. The Direct Tunnel enabled SGSNS can initiate the reactivation of the PDP context to tunnel user traffic directly from the RNC to the GGSN

or to the Serving GW introduced in Release 8. This mechanism tries to reduce the number of user-plane traffic anchors.

, keep track of the location of mobile devices and participate in GTP signaling between the GGSN and RNC.

, the Evolved Packet Core (EPC. Compared to four main GPRS PS domain entities of Release 6,

and three main functional entities in the core, i e. the Mobility Management Entity (MME), the Serving GW (S-GW) and the Packet data Network GW (PDN GW).

Release 9 (2010) introduces the definition of Home (e) Nodeb Subsystem. These systems allow unmanaged deployment of femtocells at indoor sites,

providing almost perfect broadband radio coverage in residential and working areas, and offloading the managed, pre-panned macro-cell network 14.

In Release 10 (2010) Selective IP Traffic Offload (SIPTO) and Local IP Access (LIPA) services have been published 15.

These enable local breakout of certain IP traffic from the macro-cellular network or the H (e) Nodeb subsystems,

Towards Scalable Future Internet Mobility 41 entities in the same residential/enterprise IP network without the user plane traversing the core network entities.

The above evolutionary steps resulted in that radio access networks of 3gpp became flattened to one single serving node (i e.,

the flat nature of LTE and LTE-A architectures concerns only the control plane but not the user plane:

LTE is linked to the Evolved Packet Core (EPC) in the 3gpp system evolution, and in EPC, the main packet switched core network functional entities are still remaining centralized,

keeping user IP traffic anchored. There are several schemes to eliminate the residual centralization and further extend 3gpp. 3. 2 Ultra Flat Architecture One of the most important schemes aiming to further extend 3gpp standards is the Ultra Flat Architecture (UFA) 16 20.

Authors present and evaluate an almost green field approach which is a flat and distributed convergent architecture,

with the exception of certain control functions still provided by the core. UFA represents the ultimate step toward flattening IP-based core networks

e g.,, the EPC in 3gpp. The objective of UFA design is to distribute core functions into single nodes at the edge of the network, e g.,

, the base stations. The intelligent nodes at the edge of the network are called UFA gateways.

Fig. 2 illustrates the UFA with HIP and PMIP-based mobility control. Fig. 2. The Ultra Flat Architecture with HIP and PMIP-based mobility control Since mobility introduces frequent IP-level handovers a Session Initialization Protocol (SIP) based

, the new IP ADDRESS before physical handover. This scheme supports both mobile node (MN) and network decided handovers.

Interworking with Internet applications based on non SIP control protocol is a technical challenge for mobile operators.

A Mobile IPV6 and a Host Identity Protocol (HIP) based signaling scheme alternative has been introduced for UFA by Z. Faigl et al. 18.

to reduce the number of HIP Base Exchanges in the access and core network, and to enable delegation of HIP-level signaling of the MN by the UFA GWS.

Architectures 4. 1 Motivations for Distributing Mobility Functions Flat mobile networks not only require novel architectural design paradigms, special network nodes and proprietary elements with peculiar functions

command and event services form the key routines of the future mobile Internet designs. The importance of this research area is emphasized also by the creation of a new IETF nonworking group called Distributed Mobility Management (DMM) in August 2010,

In 3g UMTS architectures centralized and hierarchical mobility anchors are Flat Architectures: Towards Scalable Future Internet Mobility 43 implemented by the RNC, SGSN and GGSN nodes that handle traffic forwarding tasks using the apparatus of GPRS Tunneling Protocol (GTP.

The similar centralization is noticeable in Mobile IP (MIP) 21 where the Home Agent an anchor node for both signaling

and tunnels user traffic towards the mobile's current locations and vice versa. Several enhancements and extensions such as Fast Handoffs for Mobile IPV6 (FMIP) 22, Hierarchical Mobile IPV6 (HMIP) 23, Multiple Care-of Addresses Registration 24

, Network Mobility (NEMO) Basic Support 25, Dual-Stack Mobile IPV6 26, and Proxy Mobile IPV6 (PMIP) 27, were proposed to optimize the performance

and broaden the capabilities of Mobile IP, but all of them preserve the centralized and anchoring nature of the original scheme.

There are also alternate schemes in the literature aiming to integrate IP-based mobility protocols into cellular architectures

Cellular IP 28 introduces a gateway router dealing with local mobility management while also supporting a number of handoff techniques and paging.

A similar approach is the handoff-aware wireless access Internet infrastructure (HAWAII) 29, which is a separate routing protocol to handle micromobility.

Some of the above solutions are standardized already 12 13 33 for 3g and beyond 3g architectures where the introduced architectural evolution is in progress:

EUTRAN (Evolved Universal Terrestrial Radio Access Network) or LTE (Long term Evolution) base stations (enodebs) became distributed in a flatter scheme allowing almost complete distribution of radio

and handover control mechanisms together with direct logical interfaces for inter-enodeb communications. Here, traffic forwarding between neighboring enodebs is allowed temporarily during handover events providing intra-domain mobility.

mobility management mechanisms in current wireless and mobile networks anchor the user traffic relatively far from users'location.

This results in centralized, unscalable data plane and control plane with non-optimal routes, overhead and high end-to-end packet delay even in case of motionless users,

but still remain in the core network. A good example is the Global HA to HA protocol 34

and distribute the Home Agents in Layer 3, at the scale of the Internet. DIMA (Distributed IP Mobility Approach) 35 can also be considered as a core-level scheme by allowing the distribution of MIP Home Agent (the normally isolated central server) to many and less powerful interworking servers called Mobility

Agents (MA. These new nodes have combined the functionality of a MIP Home Agent and HMIP/PMIP Mobility Anchor Points.

The concept of UMTS Base Station Router (BSR) 37 realizes such an access-level mobility management distribution scheme where a special network element called BSR is used to build flat cellular systems.

while a common UMTS network is built from a plethora of network nodes and is maintained in a hierarchical and centralized fashion,

Furthermore, the BSR can be considered a special wireless edge router that bridges between mobile/wireless and IP communication.

Core network nodes are mainly simple IP routers. The scheme applies DHT and Loc/ID separation:

and an IP ADDRESS based locator (Loc) changed by every single mobility event. The (Loc, ID) pair of each mobile is stored inside AGW nodes

and organized/managed using DHTS. A third type of DMM application scenarios is the so-called host-level

a special information server is required in the network, which can also Flat Architectures: Towards Scalable Future Internet Mobility 45 be centralized or distributed.

A good example for host-level schemes in the IP layer is MIPV6 which is able to bypass the user plane anchor (i e.,

, Mobile IP, NEMO BS and Proxy Mobile IP without route optimization) do not separate signaling and user planes

and data packets traverse the centralized or hierarchized mobility anchor. Since the volume of user plane traffic is compared much higher to the signaling traffic

hence separate control packets from data messages after a short period of route optimization procedure.

the algorithms supporting dynamic mobility could also be distributed. Such integration is accomplished in 44 45 where authors introduce

depending on mobiles'actual location when sessions are getting set up. The solution's dynamic nature lies in the fact that sessions of mobile nodes are anchored dynamically on different ANS depending on the IP ADDRESS used Based on this behavior,

the system is able to avoid execution of mobility management functions (e g.,, traffic encapsulation) as long as a particular mobile node is not moving.

The PMIP-based solution discusses a possible deployment scheme of Proxy Mobile IP for flat architecture.

This extension allows to dynamically distributing mobility functions among access routers: the mobility support is restricted to the access level,

both data plane and control plane are distributed). This implies the introduction of special mechanisms in order to identify the anchor that manages mobility signaling

and data forwarding of a particular mobile node, and in most cases this also requires the absolute distribution of mobility context database (e g.,

, for binding information) between every element of the distributed anchor system. Distributed Hash Table or anycast/broadcast/multicast communication can be used for the above purposes.

, by using Hi3 50 for core-level distribution of HIP signaling plane) are also feasible. 5 Conclusion Flat architectures infer high scalability

and IP-enabled radio base station (BS) entities are connected directly to the IP core infrastructure.

Towards Scalable Future Internet Mobility 47 BS nodes also minimizes the feedback time of intermodule communication, i e.,

In flat architectures the radio access network components could be compared much cheaper to HSPA and LTE devices today because of the economy of scale.

The higher competition of network management tools due to the apparition of tools developed formerly for the Internet era may reduce the operational expenditures as well.

due to lack of core controller entities base stations are managed no more centrally; hence failure diagnostics and recovery must be handled in a fully distributed and automated way.

but it comes with the benefits of scalability, fault tolerance and flexibility. Optimization of handover performance is another key challenge for flat networks.

Since all the BSS are connected directly to the IP core network, hiding mobility events from the IP layer is much harder.

and S. Imre tional hierarchical and centralized mobile telecommunication architectures. The IP network that deals with the interconnection of base stations in flat networks must be able to assure different Qos levels (e g.,

Based on the collected benefits and the actual challenges of flat architectures we can say that applying flat networking schemes together with distributed and dynamic mobility management is one of the most promising alternatives to change the current mobile Internet architecture

References 1. UMTS Forum White paper: Recognising the Promise of Mobile Broadband (June 2010) 2. Cisco VNI:

Global Mobile Data Traffic Forecast Update, 2009-2014 (Feb. 2010) 3. Dohler, M.,Watteyne, T.,Alonso-Zárate, J.:

Machine to machine-Machine: An Emerging Communication Paradigm, Tutorial. In: Globecom'10.dec 2010) 4. Schulze, H.,Mochalski, K.:

Ipoque, Internet Study 2008/2009, Ipoque (Jan. 2011) 5. UMTS Forum, REPORT NO 37, Magic Mobile Future 2010-2020 (April 2005) 6

. International Telecommunication Union, Press release: ITU sees 5 billion mobile subscriptions globally in 2010 (February 2010) 7. Cisco VNI:

Hyperconnectivity and the Approaching Zettabyte Era (June 2010) 8. ETSI GTS GSM 03.02-v5. 1. 0:

Digital cellular telecommunications system (Phase 2+)-Network architecture (GSM 03.02)( 1996) 9. 3gpp TS 23.002: Network architecture, V10. 1. 1, Release 10.jan 2011) 10. 3gpp TR 23.919:

Direct Tunnel Deployment Guideline, Release 7, V1. 0. 0 (May 2007) 11. 3gpp TS 23.401:

General Packet Radio Service (GPRS) enhancements for Evolved Universal Terrestrial Radio Access Network (E-UTRAN) access, Rel. 8, V8. 12.dec 2010) 12. 3gpp TS 29.275,

Proxy Mobile IPV6 (PMIPV6) based Mobility and Tunneling protocols, Stage 3, Release 10, V10. 0. 0 (Dec. 2010) 13. 3gpp TS 24.303

, Mobility management based on Dual-Stack Mobile IPV6, Stage 3, Release 10, V10. 1. 0 Dec (2010) 14.

Femtoforum: Femtocells Natural Solution for Offload a Femto Forum brief (June 2010) 15. 3gpp TR 23.829:

Local IP Access and Selected IP Traffic Offload, Release 10, V1. 3 (2010) 16. Daoud, K.,Herbelin, P.,Crespi, N.:

UFA: Ultra Flat Architecture for high bitrate services in mobile networks. In: Proc. of PIMRC'08, Cannes, France, pp. 1 6 (2008) 17.

Daoud, K.,Herbelin, P.,Guillouard, K.,Crespi, N.:Performance and Implementation of UFA: a SIP-based Ultra Flat Mobile network Architecture.

In: Proc. of PIMRC (Sep. 2009) 18. Faigl, Z.,Bokor, L.,Neves, P.,Pereira, R.,Daoud, K.,Herbelin, P.:

Evaluation and comparison of signaling protocol alternatives for the Ultra Flat Architecture, ICSNC, pp. 1 9 (2010) Flat Architectures:

Towards Scalable Future Internet Mobility 49 19. Bokor, L.,Faigl, Z.,Imre, S.:A Delegation-based HIP Signaling Scheme for the Ultra Flat Architecture.

Journal of Computer networks (2011), doi: 10.1016/j. comnet. 2011.02.005 21. Johnson, D.,Perkins, C.,Arkko, J.:

IP Mobility Support in IPV6, IETF RFC 3775 (2004) 22. Koodli, R. ed.:Fast Handoffs for Mobile IPV6, IETF RFC 4068 (July 2005) 23.

Soliman, H.,Castelluccia, C.,El Malki, K.,Bellier, L.:Hierarchical Mobile IPV6 Mobility Management (HMIPV6), IETF RFC 4140 (Aug. 2005) 24.

Wakikawa, R. ed.:V. Devarapalli, G. Tsirtsis, T. Ernst, K. Nagami: Multiple Care-of Addresses Registration, IETF RFC 5648 (October 2009) 25.

Mobile IPV6 Support for Dual Stack Hosts and Routers, IETF RFC 5555 (June 2009) 27.

Proxy Mobile IPV6, IETF RFC 5213 (Aug. 2008) 28. Valko: Cellular IP: A New Approach to Internet Host Mobility, ACM SIGCOMM Comp.

Commun. Rev. 29 (1), 50-65 (1999) 29. Ramjee, R.,Porta, T. L.,Thuel, S.,Varadhan, K.,Wang, S.:

Network protocols (1999) 30. Grilo, A.,Estrela, P.,Nunes, M.:Terminal Independent Mobility for IP (TIMIP.

Media Independent Handover, IEEE Std 802.21-2008 (Jan. 2009) 33. 3gpp TS 23.402, Architecture enhancements for non-3gpp accesses, Rel. 10

Global HA to HA protocol, IETF Internet-Draft, draft-thubert-nemo-global-haha-02. txt (Sept. 2006) 35.

A Distributed IP Mobility Approach for 3g SAE. In: Proc. of 19th PIMRC, ISBN: 978-1-4244-2643-0 (Sept. 2008) 36.

The UMTS base station router. Bell labs Tech. Journal, I. 11 (4), 93 111 (2007) 38.

Proc. of the 28th IEEE conference on Global telecommunications (GLOBECOM'09), Honolulu, HI (2009) 46.

Dynamic Mobile IP (DMI), IETF Internet-Draft, draft-kassi-mobileip-dmi-01. txt (Jan. 2003) 47.

Dynamic Mobility Anchoring, IETF Internet-Draft (May 2010) 49. Yan, Z.,Lei, L.,Chen, M.:

WIISE-A Completely Flat and Distributed Architecture for Future Wireless communication Systems, Wireless World Research Forum (Oct. 2008) 50.

Journal of Computer Communications 31 (10), 2457 2467 (2008) J. Domingue et al. Eds.):) Future Internet Assembly, LNCS 6656, pp. 51 66,2011.

The Author (s). This article is published with open access at Springerlink. com. Review and Designs of Federated Management in Future Internet Architectures Martín Serrano1, Steven Davy1, Martin Johnsson1, Willie Donnelly1,

and Alex Galis2 1 Waterford Institute of technology WIT Telecommunications Software and Systems Group TSSG, Co. Waterford, Ireland {jmserrano, sdavy, mjohnsson, wdonnelly}@ tssg. org

2 University college London UCL Department of Electronic and Electrical engineering, Torrington Place, London, U k. a. galis@ee. ucl. ac. uk Abstract.

The Future Internet as a design conception is network and serviceaware addressing social and economic trends in a service oriented way.

In the future Internet applications transcend disciplinary and technology boundaries following interoperable reference model (s). In this paper we discuss issues about federated management targeting information sharing capabilities for heterogeneous infrastructure.

In Future Internet architectures, service and network requirements act as design inputs particularly on information interoperability and cross-domain information sharing.

An inter-operable, extensible, reusable and manageable new Internet reference model is critical for Future Internet realisation and deployment.

We address challenges for a future Internet Architecture perspective using federation. We also provide, in a form of realistic implementations, research results

all this activities are developed under the umbrella of federated management activity in the future Internet. Keywords: Federation, Management, Reference Model, Future Internet, Architectures and Systems, Autonomics, Service Management, Semantic Modelling and Management, Knowledge Engineering, Networking Data and Ontologies, Future Communications

and Internet. 1 Introduction In recent years convergence on Internet technologies for communication's, computation's and storage's networks and services has been a clear trend in the Information and Communications technology (ICT) domain.

Although widely discussed and 52 M. Serrano et al. researched, this trend has not fully run its course in terms of implementation,

due to many complex issues involving deployment of non-interoperable and management infrastructural aspects and also due to technological,

social, economic restrictions and bottlenecks in the future Internet. In the future Internet, services and networks follow a common goal:

to provide solutions in a form of implemented interoperable mechanisms. Telecommunications networks have undergone a radical shift from a traditional circuit-switched environment with heavy/complex signalling focused on applications-oriented perspective,

towards a converged service-oriented space, mostly Internet interaction by customer as end-user and network operators as service providers.

The benefits of this shift reflect cost reduction and increase systems flexibility to react to user demands,

by replacing a plethora of proprietary hardware and software platforms with generic solutions supporting standardised development and deployment stacks.

The Future Internet as design conception is service-aware of the network infrastructure addressing service-oriented

In the future Internet trans-disciplinary solutions (applications that transcend disciplinary boundaries) following reference model (s) are crucial for a realistic integrated management realisation.

Reliable services and network performance act as technology requirements for more secure and reliable communication systems supporting end user and network requirements.

Demands on data models integration are requirements to be considered during the design and implementation phases of any ICT system.

The emergence and wide-scale deployment of wireless access network technologies calls into question the viability of basing the future Internet on IP

to rebuild the Internet, argue that the future lies in layers of overlay networks that can meet various requirements whilst keeping a very simplistic, almost unmanaged, IP for the underlying Internet.

Others initiatives such as Clean slate program 2 Stanford university, and Architecture Design Project for New Generation Network 3 argue that the importance of wireless access networks requires a more fundamental redesign of the core Internet Protocols themselves.

We argue that service agnostic network design are no longer a way to achieve interactive solutions in terms of service composition and information sharing capabilities for heterogeneous infrastructure support.

Review and Designs of Federated Management in Future Internet Architectures 53 In this paper service and network requirements 4 5 6 7 8 9 acts as inputs

particularly on information interoperability and cross-domain information sharing controlling communication systems for the Future Internet.

common and manageable new Internet reference model is critical for Future Internet realization and deployment.

The new Internet reference model must rely on the fact that high-level applications make use of diverse infrastructure representations and not use of resources directly.

We address challenges for a future Internet Architecture perspective using federation. We also provide in a form of realistic implementations, research results

Section II presents a brief review of the challenges about Future Internet architectures in terms of cross-domain interoperability.

Section III presents the rationale about federation as crucial concept in the framework of this Future Internet research.

Section VII presents the summary and outlook of this research. Finally some bibliography references supporting this research are included. 2 Challenges for Future Internet Architectures This section focuses on interdisciplinary approaches to specify data link and crossdomain interoperability to,

collectively, constitute a reference model that can guide the realisation of future communications environments in the future Internet 4 11 12 13.

The Future Internet architecture must provide societal services and, in doing so, support and sustain interactions between various communities of users in straight relation with communication infrastructure mechanisms.

Service-awareness 4 has many aspects to consider as challenges including: delivery of content and service logic with consumers'involvement and control;

interrelation and unification of the communication, storage, content and computation substrata. Networking-awareness 4 challenges imply the consumer-facing

The optimization of resources 15 16 17 using federation in the future Internet relies on classify and identify properly what resources need to be used,

and services. 3 Rationale for Federation in the future Internet Federation is relatively a new paradigm in communications,

the rationale for federated, autonomic management of communications services is addressed from the perspective of end-to-end applications and services in the future Internet.

Federation in the future Internet envisions management systems (networks and services) made up of possibly heterogeneous components, each

Future Internet environments consist of heterogeneous administrative domains, each providing a set of different services. In such complex environment, there is no single central authority;

or a set of distributed collaborating governing Review and Designs of Federated Management in Future Internet Architectures 55 authorities in

which reside on another domain are managed correctly. 4 Federated Management Activity in the future Internet This section references theoretical foundation for the development of interdisciplinary Future Internet visions about a Federated Management and their implications for networks

In future Internet end user service, application and network requirements act as guidelines to identify study and clarify part of complex requirements.

The relationships between Network Virtualisation and Federation 16 21 22 23 and the relationship between Service virtualisation (service clouds) and federation 17 are the support of a new world of solutions defining the Future Internet.

the aggregation and subsequent understanding of monitoring/fault data is a problem that has not yet been solved completely here is where federation take place

These cross-domain interactions demand certain level of abstraction to deal with mapping requirements from different information and data domains.

the software that manages them, and the actors who direct such management. In federation management end-to-end communication services involve configuring service

A goal of autonomic systems is to provide rich usage data to guide rapid service innovation.

what federation to the next generation networks and in the future internet design with service systems using heterogeneous network technologies imply.

A clear scenario where federation is being identified as useful mechanism is the Internet service provisioning

in today's Internet it is observed the growing trend for services to be provided both and consumed by loosely coupled value networks of consumers, providers and combined consumer and providers.

and Designs of Federated Management in Future Internet Architectures 57 to offer common and agreed services even with many technological restrictions

In the current Internet typical large enterprise systems contain thousands of physically distributed software components that communicate across different networks 27 to satisfy end-to-end services client requests.

, the components may be deployed in different data centres), and the diversity on service demand and network operating conditions, it is very difficult avoid conflicts 14 20 28 between different monitoring

The Figure 2 depicts the federated autonomic reference model service life cycle for the Future Internet.

and also identify particular management data at application service, middleware and hardware levels (3. Analysis) that can be gathered,

processed, aggregated and correlated (4. Mapping) to provide knowledge that will support management operations of large enterprise applications (5. Federated Agreements)

We support the idea that monitoring data at the network and application level can be used to generate knowledge that can be used to support enterprise application management in a form of control loops in the information;

a feature necessary in the future Internet service provisioning process (7. Federated Decisions. Thus infrastructure can be re-configurable

and resolve high level representation and mapping of data and information. Negotiations in form of data representation between different data and information models by components in the system (s) are associated to this feature.

Management Control-Administration functionality for the establishment of cross-domain regulations considering service and network regulations and requirements as negotiations.

Review and Designs of Federated Management in Future Internet Architectures 59 5 Federated Management Architecture This section describes designing principles for inter-domain federated management architectures

in the Future Internet. These designs about architecture for the federated reference model by functional blocks addresses the specification of mechanisms including models, algorithms, processes, methodologies and architectures.

The functional architecture collectively constitute, in terms of implementation efforts, framework (s), toolkit (s) and components that can guide the realisation of federated communications environments to effectively provide complex services (interoperable boundaries) and,

and sustain service offering between various communities of users (heterogeneous data & infrastructure). The federated architecture must be enabled for ensuring the information is available allowing useful transfer of knowledge (information interoperability) across multiple interfaces.

and configurations for managing services and networks are used to ensure transference of results to other systems as result of sensitivity analysis.

information and data can be integrated, and the power of machinebased learning and reasoning can be exploited more fully.

Autonomic control loops and its formalisms 29 30, such as FOCALE 25 and Autoi 21 23 translate data from a device-specific form to a device

In a federated autonomic architecture, information is used to relate knowledge, rather than only map data,

and described by the man Review and Designs of Federated Management in Future Internet Architectures 61 agement distribution.

apartment buildings, offices) generates more demand in deploying wireless 802.11-based mesh networks this expansion will be a patchwork of mesh networks;

62 M. Serrano et al. 6. 2 Federation of Network and Enterprise Management Systems Typical large enterprise systems contain thousands of physically distributed software components that communicate across different networks

Challenges in this scenario relies on how monitoring at the network level can provide knowledge that will enable enterprise application management systems to reconfigure software components to better adapt applications to prevailing network conditions.

Within the Web 2. 0 development online value is expanding from searching and e-consumerism applications,

to participative applications including blogs, wikis, online social networks, RSS feeds, Instant Messaging, P2p applications, online gaming and increasingly pervasive Voip applications.

Value networks share with Web 2. 0 application users a concern with value of interacting effectively with rest of the network community (federation.

such as recommending different network or service providers to their mem Review and Designs of Federated Management in Future Internet Architectures 63 bers.

Achieving this requires increased degrees of integration between telecommunications network management systems and devices. In particular, it is important to develop methods (management functions) through

and correlation techniques that can process relevant data in a timely and decentralised manner and relay it as appropriate to management federated making functions are necessaries to investigate (federation).

and Outlook In the future Internet new designs ideas of Federated Management in Future Internet Architectures must consider high demands of information interoperability to satisfy service composition requirements being controlled by diverse,

The federated autonomic reference model approach introduced in this paper as a design practice for Future Internet architectures emerges as an alternative to address this complex problem in the future Internet of networks and services.

We have studied how federation brings support for realisation on the investigated solution (s) for information interoperability and cross-domain information sharing controlling communication systems in the Future Internet.

Algorithms and processes to allow federation in enterprise application systems to visualize software components, functionality and performance.

Techniques for analysis, filtering, detection and comprehension of monitoring data in federated enterprise and networks.

Algorithms and processes to allow federated application management systems reconfigure or redeploy software components realizing autonomic application functionality.

Guidelines and exemplars for the exchange of relevant knowledge between network and enterprise application management systems.

This paper makes references to design foundations for the development of federated autonomic management in architectures in the future Internet.

References 1. NSF-funded initiative to rebuild the Internet (Online: Oct. 2010), http://www. geni. net/2. Clean slate Program, Stanford university (Online:

Oct. 2010), http://akari-project. nict. go. jp/eng/index2. htm Review and Designs of Federated Management in Future Internet Architectures 65 4. Galis

Management and Service-aware Networking Architectures (MANA) for Future Internet Position Paper: System Functions, Capabilities and Requirements (Invited paper.

Future Generation Internet Architecture. Newarch Final Technical Report, http://www. isi. edu/newarch/6. van der Meer, S.,Davy, A.,Davy, S.,Carroll, R.,Jennings, B

From Autonomic Computing to Autonomic Networking: an Architectural Perspective. In: Proc. of 5th IEEE Workshop on Engineering of Autonomic and Autonomous Systems (EASE 2008),

Rethinking the design of the Internet: the end to end arguments vs. the brave new world. ACM Transactions on Internet Technology 1 (1)( 2001) 11.

Subharthi, P.,Jianli, P.,Raj, J.:Architectures for the Future Networks and The next Generation Internet:

A Survey. Computer Communications (July 2010), 63 pp. http://www1. cse. wustl. edu/jain/papers/ftp/i3survey. pdf 12.

Curran, K.,Mulvenna, M.,Galis, A.,Nugent, C.:Challenges and Research Directions in Autonomic Communications.

International Journal of Internet Protocol Technology (IJIPT) 2 (1)( 2006) 13. Rubio-Loyola, J.,Astorga, A.,Serrat, J.,Chai, W. K.,Mamatas, L.,Galis, A.,Clayman, S.,Cheniour, A.,Lefevre, L.,Fischer, A.,Paler, A.,Al

Platforms and Software systems for an Autonomic Internet. In: IEEE Globecom 2010, Miami, USA, 6-10 december (2010) 14.

Challenges for Federated, Autonomic Network Management in the future Internet. In: Manfi workshop, June, NY, USA (2009) 15.

An Inference Plane to Support The next Generation Internet. In: IEEE GIIS 2007,2-6 july (2007) 17.

A Formal Approach for the Inference Plane Supporting Integrated Management Tasks in the future Internet. In: 1st IFIP/IEEE Manfi Intl Workshop, In conjunction 11th IEEE IM2009, Long island, NY, USA, June 2009, IEEE Computer Society Press, Los

Alamitos (2009) 19. Brennan, R.,Feeney, K.,Keeney, J.,O'sullivan, D.,Fleck II, J.,Foley, S.,van der Meer, S.:

Federation, A Matter of Autonomic Management in the future internet. In: IEEE/IFIP Network Operations & Management Symposium, NOMS 2010, Osaka, Japan, 19-23 april (2010) 21.

Autonomic Internet: A Perspective for Future Internet Services Based on Autonomic Principles. In: 2007 IEEE Management Week Manweek 2007 2nd IEEE MACE 2007 Workshop, San José, CA, USA, 29.oct 2 Nov (2007) 22.

Rochwerger, B.,et al.:An Architecture for Federated Cloud computing. In: Cloud computing: Principles and Paradigms, Wiley, ISBN:

0470887990 (April 2011) 23. Galis, A.,et al.:Management Architecture and Systems for Future Internet Networks.

In: Towards the Future Internet A European Research Perspective, p. 350. IOS Press, Amsterdam (2009) 24.

Feldmann, A.:Internet clean-slate design: what and why? ACM SIGCOM Computer Communication Review 37 (3)( 2007) 25.

Strassner, J.,Agoulmine, N.,Lehtihet, E.:FOCALE A Novel Autonomic Networking Architecture. ITSSA Journal 3 (1), 64 79 (2007) 26.

Foley, S n.,Zhou, H.:Authorisation Subterfuge by Delegation in Decentralised Networks. In: Proc. of the 13th International security Protocols Workshop, Cambridge, UK (April 2005) 27.

Jennings, B.,et al.:Towards Autonomic Management of Communications networks. IEEE Comms Magazine 45 (10), 112 121 (2007) 28.

Strassner, J.:Autonomic Networks and Systems: Theory and Practice. In: NOMS 2008 Tutorial, Brasil (April 2008) 29.

Management and Context Integration Based on Ontologies for Pervasive Service Operations in Autonomic Communication systems. Phd Thesis, UPC (2008) 31.

) Future Internet Assembly, LNCS 6656, pp. 67 80,2011. The Author (s). This article is published with open access at Springerlink. com. An Architectural Blueprint for a Real-world Internet Alex Gluhak1, Manfred Hauswirth2, Srdjan Krco3, Nenad Stojanovic4, Martin

Bauer5, Rasmus Nielsen6, Stephan Haller5, Neeli Prasad6, Vinny Reynolds2, and Oscar Corcho8 1 University of Surrey, UK 2 National University of Galway, Ireland 3 Ericsson, Serbia 4 FZI, Germany 5 NEC, Germany

Numerous projects in the area of Real-world Internet (RWI Internet of things (Iot), and Internet Connected Objects have proposed architectures for the systems they develop.

All of these systems are faced with very similar problems in their architecture and design and interoperability among these systems is limited.

To address these issues and to speed up development and deployment while at the same time reduce development and maintenance costs,

Real-world Internet, Internet of things, Internet Connected Objects, Architecture 1 Introduction Devices and technologies ubiquitously deployed at the edges of the networks will provide an infrastructure that enables augmentation of the physical world and interaction with it, without the need for direct human intervention,

thus creating the essential foundations for the Real-world Internet (RWI). Leveraging the collective effort of several projects over the last number of years SENSEI, ASPIRE, IOT-A, PECES, CONET, SPITFIRE, Semsorgrid4env,

this chapter presents the current status of the work aimed at definition of an RWI reference architecture.

The core contribution of this paper is the distillation of an initial model for RWI based on an analysis of these state of art architectures and an understanding of the challenges.

An identification of a core set of functions and underlying information models, operations and interactions that these architecture have in common.

and models and what features they provide. 68 A. Gluhak et al. 2 The Real world Internet Since the introduction of the terminology over a decade ago,

the"Internet of things (Iot)" has undergone an evolution of the underlying concepts as more and more relevant technologies are maturing.

which all physical objects are tagged by Radio frequency identification (RFID) transponders in order to be identified uniquely by information systems. However the concept has grown into multiple dimensions,

encompassing sensor networks able to provide real world intelligence or the goal-oriented autonomous collaboration of distributed objects via local wireless networks or global interconnections such as the Internet.

Kevin Ashton, former Director of the Auto-ID Center, once famously formulated: Adding radiofrequency identification

and other sensors to everyday objects will create an Internet of things, and lay the foundations of a new age of machine perception.

We believe that machine perception of the real world is still at the heart of the Internet of things, no matter

As such, one of the key roles of the Internet of things is to bridge the physical world

and its representation in the digital world of information systems, enabling what we refer to in part of the Future Internet Assembly (FIA) community as the so called Real world Internet (RWI).

The RWI is the part of a Future Internet that builds upon the resources provided by the devices HAL of the Internet of things, offering real world information and interaction capabilities to machines,

software artifacts and humans connected to it. The RWI assumes that the information flow to

and from Iot devices is taking place via local wired and wireless communication links between devices in their proximity and/or through global interconnections in the form of the current Internet and mobile networks or future fixed and mobile network infrastructures.

One important property of the RWI which distinguishes it from the current Internet is its heterogeneity

both regarding the types of devices as well as communication protocols used. IPV6 and in particular 6lowpan play an important role,

but other proprietary wireless protocols will see continued use as well. To deal with this heterogeneity, services in the form of standard Web Services and DPWS1,

but more likely using RESTFUL approaches and application protocols like Coap provide a useful abstraction.

As services play a pivotal role in the future Internet Architecture, the use of services for integrating the RWI also fits well into the overall architectural picture.

One has to keep in mind though that RWI services have some different properties from common

enterprise-level services: They are of lower granularity, e g.,, just providing simple sensor readings and, more importantly,

and the data they deliver has to be associated with some quality of information parameters before further processing. 1 Device Profile for Web Services An Architectural Blueprint for a Real-world Internet 69 3 Reference Architecture In this section we present an initial model on

1. Underlying system assumptions, 2. functional coverage of the services provided by the architectures, 3. underlying information models in the architectures,

in order to monitor and interact with the physical entities that we are interested in. The digital world consists of:

or application software that intends to interact with Resources and Eoi. Providing the services and corresponding underlying information models to bridge the physical

Entity Level Resource Level Real world sensor RFID actuator sensor sensor Entity-based Context Model models relevant aspects of Real world Real-world Internet

or architectural services. 3. 1 Functional Coverage of RWI Architectures This section explores the different functional features provided by the service functions of the existing architectures to support the interactions between resources and resource users and the corresponding

It enables the dy An Architectural Blueprint for a Real-world Internet 71 namic instantiation of resources (e g.,

Accountability and traceability can be achieved by recording transactions and interactions taking place at the respective system entities. 3. 2 Smart Object model At its core,

Conceptually, resources provide unifying abstractions for real-world information and interaction capabilities comparable to web resources in the current web architecture.

In the same way as a web user interacts with a web resource, e g.,, retrieve a web page,

the user can interact with the real-world resources, e g.,, retrieve sensor data from a sensor.

However, while the concept of the web resource refers to a virtual resource identified by a Universal Resource Identifier (URI),

a resource in the RWI context is an abstraction for a specific set of physical and virtual resources.

actuation, processing of context and sensor data or actuation loops, and management information concerning sensor/actuator nodes, gateway devices or entire collections of those.

and the software components implementing the interaction endpoints from the user perspective (Resource End point REP). Furthermore,

and their relationships in the RWI system model A REP is a software component that represents an interaction end-point for a physical resource.

In comparison to the current web architecture, REPS can be considered equivalent to web resources, which are identified uniquely by a URI.

mobile phones or access points that embed resources. A REP Host is a device that executes the software process representing the REP. As mentioned before,

the resources and REPS are separated conceptually from their hosts to facilitate different deployment options. In some cases a REP host and a resource host can be located co on the same physical device

, in the case of a mobile phone. Similarly, there may be cases where the REP is not hosted on the resource host itself, for example,

a computer in the network or an embedded server may act as the REP host for a resource,

low-power sensor nodes, from attacks by hosting their REPS on more powerful hardware. Unlike other models, the Smart Object model considers also real-world entities in its model

or objects of the real world that are considered relevant to provide a service to An Architectural Blueprint for a Real-world Internet 73 users or applications.

n 74 A. Gluhak et al. 4 Analysis of Existing Architectures In this section we briefly review five of the most relevant RWI architecture approaches with respect to the functional coverage provided in the context of the above defined reference architecture.

and a table at the section's end summarizes the functional coverage of the five main architectures. 4. 1 ASPIRE The ASPIRE architecture ASPIRE is based on EPGGLOBAL EPC with a number of objective-specific additions.

In a Radio frequency identification (RFID) based scenario the tags act as hosts for the resources in form of Electronic Product Codes (EPCS), IDS or other information as well as for value-added information in form of e g. sensor data.

The resource hosts are abstracted through the RFID readers due to the passive communication of the tags.

The Object Naming Service (ONS) corresponds to the Entity Directory that returns the URLS of relevant resources for the EPC in question this is the White Pages service.

ASPIRE introduces a Business Event Generator (BEG) which implements additional logic for interactions using semantics of the specific RFID application.

The system is based on the OSGI service middleware and consists of two main sub systems: the service platform openaal and the ETALIS Complex event processing system (icep. fzi. de.

It provides generic platform An Architectural Blueprint for a Real-world Internet 75 services like context management for collecting

and abstracting data about the environment, workflow based specifications of system behaviour and semanticallyenabled service discovery.

i e. eventdriven one. 4. 3 PECES The PECES architecture PECES provides a comprehensive software layer to enable the seamless cooperation of embedded devices across various smart spaces on a global scale in a context-dependent

The PECES middleware architecture enables dynamic group-based communication between PECES applications (Resources) by utilizing contextual information based on a flexible context ontology.

Although Resources are not directly analogous to PECES middleware instances, gateways to these devices are more resource-rich

and can host middleware instances, and can be queried provided that an application-level querying interface is implemented.

must be running the PECES middleware before any interaction may occur. Both one-shot and continuous interactions are supported between components

streaming and static data sources in manners that were not necessarily foreseen when the sensor networks were deployed

or the data sources made available. The architecture may be applied to almost any type of real world entity,

streaming data sources, normally containing historical information from sensors; and even relational databases, which may contain any type of information from the digital world (hence resource hosts are multiple).

These resources are made available through a number of data-focused services (acting as resource endpoints),

which are based on the WS-DAI specification for data access and integration and which are supported by the Semsor-Grid4env reference implementation.

These services include those focused on data registration and discovery (where a spatiotemporal extension of SPARQL stsparql-,is used to discover data sources from the Semsorgrid4env registry),

data access and query (where ontology-based and non-ontology-based query languages are provided to access data:

SPARQL-Stream and SNEEQL a declarative continuous query language over acquisition sensor networks, continuous streaming data,

and traditional stored data), and data integration (where the ontology-based SPARQL-Stream language is used to integrate data from heterogeneous and multimodal data sources).

Other capabilities offered by the architecture are related to supporting synchronous and asynchronous access modes, with subscription/pull

and push-based capabilities, and actuating over sensor networks, by in-network query processing mechanisms that take declarative queries

services and resources. 4. 5 SENSEI The SENSEI architecture SENSEI aims at integrating geographically dispersed and internet interconnected heterogeneous WSAN (Wireless Sensor and Actuator Networks) systems into a homogeneous

which is inspired strongly by service oriented principles and semantic web technologies. In the SENSEI architecture each real world resource is described by a uniform resource description,

the An Architectural Blueprint for a Real-world Internet 77 architecture provides a semantic query support,

and Iot-A Iot-A as these projects have started just and have not produced architectures yet, they can only be included in the future work on an RWI reference architecture.

SPITFIRE aims at extending the Web into the embedded world to form a Web of Things (Wot),

where Web representations of real-world entities offer services to access and modify their physical state

and to mash up these real-world services with traditional services and data available in the Web.

supporting heterogeneous and resourceconstrained devices, its extensive use of existing Web standards such as RESTFUL interfaces and Linked Open Data,

The Iot-A project extends the concepts developed in SENSEI further to provide a unified architecture for an Internet of things.

It aims at the creation of a common architectural framework making a diversity of real world information sources such as wireless sensor networks and heterogeneous identification technologies accessible on a Future Internet.

which a future Iot architecture will be based, such as a global resolution infrastructure that allows Iot resources to be resolved dynamically to entities of the real world to

which they can relate. 78 A. Gluhak et al. 4. 7 Summary of Project Realizations Table 2a.

Functional coverage of current RWI architecture approaches LLAAL Using an RDF-based registry Contextual Manager provides an ontologybased information storage that captures sensor information

and selects and combines those services to achieve the (abstract) service goals Semsorgrid4env Using an RDF-based registry of data sources,

and corresponding stsparql queries In-network query processing capabilities (SNEE) with mote-based sensor networks Data services are generated dynamically according to WS-DAI (Web Services Data Access and Integration) indirect

or federated (peered) resource directory as a rendezvous point that stores resource descriptions SPARQL based query interface,

resource creation An Architectural Blueprint for a Real-world Internet 79 Table 2b. Functional coverage of current RWI architecture approaches LLAAL N/A n/A n/A LL AAL Sensor-level ontology.

It supports integration of sensors and AAL services Context Ontology: low-and top-level. It supports context reasoning from a low-level sensor-based model to a high-level service-oriented model Semsorgrid4env Limited management, through WS-DAI indirect access mode N/A n

/A According to W3c Semantic Sensor Network Ontology Observation&measuremen t, role, agent, service and resource ontologies PECES Implicit via middleware Expressive (based on ontologies),

Role-based access control for individual middleware components N/A EPC and value-added sensing EPCIS standard SENSEI Execution manager responsible for maintenance of long lasting requests

The work on the Iot reference architecture will continue to be driven by the RWI group of the FIA in collaboration with the FP7 IOT-i coordinated action project (http://www. iot-i. eu) and the IERC

the European Research Cluster on the Internet of things (http://www. internet-of-things-research. eu/).The results will be contributed to the FIA Architecture track.

References ASPIRE Advanced Sensors and lightweight Programmable middleware for Innovative RFID Enterprise applications, FP7, http://www. fp7-aspire. eu/CONET Cooperating Objects Noe,

The Things in the Internet of things. Poster at the Internet of things Conference, Tokyo (Iot, 2010)( 2010), available at http://www. iot-a. eu/public/news/resources/Thethingsintheinternetof Things sh. pdf Accessed Jan 24, 2011

Iot-A EU FP7 Internet of things Architecture project, http://www. iot-a. eu/public LLAAL FZI Living Lab AAL, http://aal. fzi. de

/PECES PERVASIVE Computing in Embedded systems, FP7, http://www. ict-peces. eu/Semsorgrid4env Semantic Sensor Grids for Rapid Application Development for Environmental Management, FP7

, http://www. semsorgrid4env. eu/SENSEI Integrating the Physical with the Digital World of the Network of the Future

FP7, http://www. ict-sensei. org SPITFIRE Semantic-Service Provisioning for the Internet of things using Future Internet Research By experimentation, FP7, http://www. spitfireproject. eu/ZGL

From Today's INTRANET of Things to a Future INTERNET OF THINGS: A Wireless-and Mobility-Related View.

IEEE Wireless communications 17 (6)( 2010) J. Domingue et al. Eds.):) Future Internet Assembly, LNCS 6656, pp. 81 90,2011.

The Author (s). This article is published with open access at Springerlink. com. Towards a RESTFUL Architecture for Managing a Global Distributed Interlinked Data-Content-Information Space Maria Chiara Pettenati, Lucia

Ciofi, Franco Pirri, and Dino Giuli Electronics and Telecommunications Department, University of Florence, Via Santa marta, 3 50139 Florence, Italy {mariachiara. pettenati, lucia. ciofi, franco. pirri, dino

. giuli}@ unifi. it Abstract. The current debate around the future of the Internet has brought to front the concept of Content-Centric architecture, lying between the Web of Documents and the generalized Web of Data

in which explicit data are embedded in structured documents enabling the consistent support for the direct manipulation of information fragments.

In this paper we present the Interdatanet (IDN) infrastructure technology designed to allow the RESTFUL management of interlinked information resources structured around documents.

IDN deals with globally identified, addressable and reusable information fragments; it adopts an URI-based addressing scheme;

uniform Web-based interface to distributed heterogeneous information management; it endows information fragments with collaboration-oriented properties, namely:

Web of Data; future Web; Linked Data; RESTFUL; read-write Web; collaboration. 1 Introduction There are many evolutionary approaches of the Internet architecture

which are at the heart of the discussions both in the scientific and industrial contexts:

Web of Data/Linked Data, Semantic web, REST architecture, Internet of Services, SOA and Web Services and Internet of things approaches.

Each of these approaches focus on specific aspects and objectives which underlie the high level requirements of being a driver towards a better Internet or a better Web.

Three powerful concepts present themselves as main drivers of the Future Internet 1 2. They are:

a user-centric perspective, a service-centric perspective and a contentcentric perspective. The user-centric perspective emphasizes the end-user experience as the driving force for all technological innovation;

the service-centric perspective is influenced currently in enterprise IT environment and in the Web2. 0 mashup culture, showing the importance of flexibly reusing service components to build efficient applications.

The Content-Centric perspective leverages on the importance of creating, pub 82 M. C. Pettenati et al. lishing and interlinking content on the Web and providing content-specific infrastructural services for (rich media

) content production, publication, interlinking and consumption. Even if it is very difficult to provide a strict separation of approaches

Table 1. Rough classification of main driving forces in current Future Network evolutionary approaches Content-centric Service-centric Users-centric Approaches Web of Data

/Linked Data REST Internet of Services WS-*SOA Web 2. 0, Web 3. 0, Semantic web Internet of things The three views can be interpreted as emphasizing different aspect rather than expressing opposing statements.

Hence, merging and homogenizing towards an encompassing perspective may help towards the right decision choice for the Future Internet.

Such an encompassing perspective has been discussed in terms of high-level general architecture in 1 and has been named Content-Centric Internet.

At the heart of this architecture is the notion of Content, defined as any type

content, service-oriented) the Future Internet Architecture herewith described essentially proposes a Virtual Resources abstraction required for the Content-Centric approach.

Another view of Content-centric Internet architecture is elaborated in 2 by Danny Ayers, based on the assumption that

therefore a Transitional Web lying between the Web of Documents and the generalized Web of Data in

which explicit data are embedded in documents enabling the consistent support for the direct manipulation of information as data without the limitation of current data manipulation approaches.

Abstracting from the different use of terms related to the concepts data, content and information which can be found in literature with different meanings 4,

the grounding consistency that can be highlighted is need related to the of providing an evolutionary direction to the network architecture hinging on the concept of a small, Web-wide addressable data/content/information unit

and handled by the network architecture so as to Managing a Global Distributed Interlinked Data-Content-Information Space 83 provide basic Services at an infrastructural level

Among the different paths to the Web of Data the one most explored is adding explicit data to content.

Directly treating content as data has had instead little analysis. In this paper we discuss evolution of Interdatanet (IDN) an high-level Resource Oriented Architecture proposed to enable the Future Internet approaches (see 5 6

and references therein). Interdatanet is composed of two main elements: the IDN-Information Model and the IDN-Service Architecture.

which different actors collaborate 3. the infrastructural support to collaboration on documents and their composing information fragments 4. the Web-wide scalability of the approach.

The purpose of this paper is to show that Interdatanet can provide a high-level model of the Content-Centric Virtualized Network grounding the Future Internet Architecture.

as highlighted in Figure 1. Fig. 1. Interdatanet architecture situated with respect to the Future Internet architecture envisaged in 7. 84 M. C. Pettenati et al.

though aiming at dealing with distributed granular content over the Web, suffer from a main limitation:

the more we get away from the data and move into the direction of information, the fewer available solutions are there capable of covering the following requirements:

(i e. non application-dependent) support to collaboration on above documents and their composing information fragments-the uniform REST interaction with the resources-the Web-wide scalability of the approach.

addressable and reusable information fragments (as in Web of Data) 2. IDN adopts an URI-based addressing scheme (as in Linked Data) 3. IDN provides simple a uniform Web-based

interface to distributed heterogeneous data management (REST approach) 4. IDN provides-at an infrastructural level-collaboration-oriented basic services, namely:

This will alleviate application-levels of sharing arbitrary pieces of information in ad hoc manner while providing compliancy with current network architectures and approaches such as Linked Data, RESTFUL Web Services, Internet of Service,

Internet of things. 2. 1 The Interdatanet Information Model and Service Architecture IDN framework is described through the ensemble of concepts,

Managing a Global Distributed Interlinked Data-Content-Information Space 85 IDN-SA (Interdatanet Service Architecture.

The IDN-SA exposes an IDN-API (Application programming interface) on top of which IDN-compliant Applications can be developed.

The Information Model is based the graph data model (see Figure 3) to describe interlinked data representing a generic document model in IDN

Generic information modeled in IDN-IM is formalized as an aggregation of data units. Each data unit is assigned at least with a global identifier

and contains generic data and metadata; at a formal level, such data unit is a node in a Directed Acyclic Graph (DAG.

The abstract data structure is named IDN-Node. An IDN-Node is the content-item handled by the content-centric IDN-Service Architecture.

The degree of atomicity of the IDN Nodes is related to the most elementary information fragment

whose management is needed in a given application. The information fragment to be handled in IDN-IM compliant documents,

An IDN-document structures data units is composed of nodes related to each other through directed links. Three main link types are defined in the Information Model:

-back links, to enable the mechanism of notification of IDN-nodes updates to parentnodes (back propagation.

Managing a Global Distributed Interlinked Data-Content-Information Space 87 Replica Management (RM) provides a delocalized view of the resources to the upper layer.

univocally and persistently identify the resources within IDN-middleware independent of their physical locations; in the lower layer are used Uniform Resource Locators (URL) to identify resource replicas as well as to access them.

The implementations of IDN-SA are a set of different software modules one module for each layer.

Each module, implemented using an HTTP server, will offers a REST interface. The interaction between IDN-compliant applications and IDN-SA follows the HTTP protocol as defined in REST architectural style too.

therefore be enabled to the manipulation of data on a global scale within the Web. REST interface has been adopted in IDN-SA implementation as the actions allowed on IDN-IM can be translated in CRUD style operations over IDN-Nodes with the assumption that an IDN-document can be thought as an IDN-Node resources collection.

which are coded in an IDN/XML format (data format defined with XML language). Every resource in such format must be well formed with respect to XML syntax,

and deploy specific functionalities of the architecture Fig. 5. Interdatanet Service Architecture scalability features Managing a Global Distributed Interlinked Data-Content-Information Space 89 without the need to achieve the complete

The presented approach is not an alternative to current Web of Data and Linked Data approaches rather it aims at viewing the same data handled by the current Web of Data from a different perspective,

where a simplified information model, representing only information resources, is adopted and where the attention is focused on collaboration around documents

or suggesting new methods of handling data, relying on standard Web techniques. Interdatanet could be considered to enable a step ahead from the Web of Document

and possibly grounding the Web of Data, where an automated mapping of IDNIM serialization into RDF world is made possible using the Named Graph approach 9. Details on this issue are beyond the scope of the present paper.

The authors are aware that the IDN vision must be confronted with the evaluation of the proposed approach.

Providing demonstrable contribution to such a high level goal is not an easy task, as it is demonstrated by the state of the art work defending this concept idea which,

b) using HTTP URIS to address information fragments to manage resources in as well as on the Web 11;

c) the adoption of a RESTFUL Web Services, also known as ROA Resource Oriented Architecture to leverage on REST simplicity (use of well-known standards i e.

The implemented Web application allows Public Officers to assess current citizens'official residence address requesting certificates to the entitled body,

because it offers infrastructural enablers to Web-based interoperation without requiring major preliminary agreements between interoperating parties

thus providing a contribution in the direction of taking full advantage of the Web of Data potential.

Towards a Content-Centric Internet. In: Tselentis, G.,Galis, A.,Gavras, A.,Krco, S.,Lotz, V.,Simperl, E.,Stiller, B.,Zahariadis, T. eds.

Towards the Future Internet-Emerging Trends from European Research, pp. 227 236. IOS Press, Amsterdam (2010) 2. Ayers, D.:

From here to There. IEEE Internet Comput 11 (1), 85 89 (2007) 3. European commission Information Society and Media.

Future Networks The way ahead! European communities: Belgium (2009) 4. Melnik, S.,Decker, S.:A Layered Approach to Information Modeling and Interoperability on the Web.

In: Proceedings ECDL'00 Workshop on the Semantic web, Lisbon (September 2000) 5. Pettenati, M. C.,Innocenti, S.,Chini, D.,Parlanti, D.,Pirri, F

. 2008) Interdatanet: A Data Web Foundation For The Semantic web Vision. Iadis International Journal On Www/Internet 6 (2 december 2008) 6. Pirri, F.,Pettenati, M. C.,Innocenti, S.,Chini, D.,Ciofi, L.:

Interdatanet: a Scalable Middleware Infrastructure for Smart Data Integration, in D. In: Giusto, D.,et al.

eds.)) The Internet of things: 20th Tyrrhenian Workshop on Digital communications, Springer, Heidelberg (2009), doi: 10.1007/978-1-4419-1674-7 12 7. Zahariadis, T.,Daras, P.,Bouwen, J.,Niebert, N.,Griffin, D.,Alvarez, F.,Camarillo, G.:

Towards a Content-Centric Internet Plenary Keynote address. Presented at Future Internet Assembly (FIA) Valencia, SP, 15-16 april (2010) 8. Richardson, L.,Ruby, S.:

RESTFUL Web Services; O'reilly Media, Inc.:Sebastopol, CA, USA (2007) 9. Carroll, J. J.,Bizer, C.,Hayes, P.,Stickler, P.:

Named graphs, provenance and trust. In: Proceedings of the 14th international conference on World wide web-WWW'05.

Presented at the 14th international conference, Chiba, Japan, p. 613. Chiba, Japan (2005), doi: 10.1145/1060745.1060835 10.

Zweben, S. H.,Edwards, S. H.,Weide, B. W.,Hollingsworth, J. E.:The Effects of Layering and Encapsulation on Software Development Cost and Quality.

IEEE Trans. Softw. Eng. 21 (3), 200 208 (1995) 11. Hausenblas, M.:Web of Data.

Oh it is data on the Web posted on April 14, 2010; accessed September 8, 2010, http://webofdata. wordpress. com/2010/04/14/ohit-is-data-on-the-web/J. Domingue et al.

Eds.):) Future Internet Assembly, LNCS 6656, pp. 91 102,2011. The Author (s). This article is published with open access at Springerlink. com. A Cognitive Future Internet Architecture Marco Castrucci1, Francesco Delli Priscoli1, Antonio Pietrabissa1,

and Vincenzo Suraci2 1 University of Rome La Sapienza, Computer and System Sciences Department Via Ariosto 25,00185 Rome, Italy {castrucci, dellipriscoli, pietrabissa}@ dis

. uniroma1. it 2 Università degli studi e-Campus Via Isimbardi 10,22060 Novedrate (CO), Italy vincenzo. suraci@uniecampus. it Abstract.

This Chapter proposes a novel Cognitive Framework as reference architecture for the Future Internet (FI),

which is based on so-called Cognitive Managers. The objective of the proposed architecture is twofold. On one hand

it aims at achieving a full interoperation among the different entities constituting the ICT environment, by means of the introduction of Semantic Virtualization Enablers,

in charge of virtualizing the heterogeneous entities interfacing the FI framework. On the other hand, it aims at achieving an inter-network and inter-layer cross-optimization by means of a set of so-called Cognitive Enablers,

which are in charge of taking consistent and coordinated decisions according to a fully cognitive approach, availing of information coming from both the transport and the service/content layers of all networks.

Future Internet architecture, Cognitive networks, Virtualization, Interoperation. 1 Introduction Already in 2005, there was the feeling that the architecture

and protocols of the Internet needed to be rethought to avoid Internet collapse 1. However,

the research on Future Internet became a priority only in the last five years, when the exponential growth of small and/or mobile devices and sensors, of services and of security requirements began to show that current Internet is becoming itself a bottleneck.

Two main approach have been suggested and investigated: the radical approach 2, aimed at completely redesign the Internet architecture,

and the evolutionary approach 3, trying to smoothly add new functionalities to the current Internet towards.

Right now, the technology evolution managed to cover the lacks of current Internet architecture, but, probably, the growth in Internet-aware devices and the always more demanding requirements of new services and applications will require radical architecture enhancements very soon.

This statement is proved by the number of financed projects both in the USA and in Europe. 92 M. Castrucci et al.

In the USA, there are significant initiatives. Nets 4 (Networking Technology and Systems) was a program of the National Science Foundation (NSF) on networking research with the objectives of developing the technology advances required to build next generation networks

GENI 6 (Global Environment for Network Innovations) is a virtual laboratory for at scale experimentation of network science, based on a 40 Gbps real infrastructure.

In Europe, Future Internet research has been included as one of the topics in FP6 and FP7.

and applications by utilizing the current Internet infrastructure. For instance, G-Lab 8 (Design and experiment the network of the future, Germany), is the German national platform for Future Internet studies,

includes both research studies of Future Internet technologies and the design and setup of experimental facilities.

GRIF 9 (Research Group for the Future Internet, France) and Internet del Futuro 10 (Spain) promotes cooperation based on several application areas (e g.,

, health) and technology platforms. FIRE 11 is an EU initiative aimed at the creation of an European Experimental Facility

which is constructed by progressively connecting existing and upcoming testbeds for Future Internet technologies. The contribution of this Chapter is the proposal of a Future Internet architecture

which seamlessly cope with the evolutionary approach but is also open to innovative technologies and services. The main idea is to collect

, users, contents, services, network resources, computing resources, device characteristics) via virtualization and data mining functionalities; the metadata produced in this way are then input of intelligent cognitive modules

Section 3 describes the Future Internet platform in detail; experimental results showing the potential of the platform are described in Section 4;

finally, Section 5 draws the conclusions. 2 Architecture Concept A more specific definition of the entities involved in the future Internet,

as well as of the Future Internet target, can be as follows: Actors represent the entities whose requirement fulfillment is the goal of the Future Internet;

for instance, Actors include users, developers, network providers, service providers, content providers, etc.;A Cognitive Future Internet Architecture 93 Resources represent the entities that can be exploited for fulfilling the Actors'requirements;

example of Resources include services, contents, terminals, devices, middleware functionalities, storage, computational, connectivity and networking capabilities, etc.;

Applications are utilized by the Actors to fulfill their requirements and needs exploiting the available resources.

In the authors'vision, the Future Internet target is to allow Applications to transparently, efficiently and flexibly exploit the available Resources,

the Future Internet should overcome the following main limitations. i) A first limitation is inherent in the traditional layering architecture

which forces to keep algorithms and procedures, laying at different layers, independent one another. In addition, even in the framework of a given layer, algorithms and procedures dealing with different tasks are designed often independently one another.

These issues greatly simplify the overall design of the telecommunication networks and greatly reduce processing capabilities,

since the overall problem of controlling the telecommunication network is decoupled in a certain number of much simpler sub-problems.

Nevertheless, a major limitation of this approach derives from the fact that algorithms and procedures are poorly coordinated one another,

impairing the efficiency of the overall telecommunication network control. The issues above claim for a stronger coordination between algorithms and procedures dealing with different tasks.

ii) A second limitation derives from the fact that, at present, most of the algorithms and procedures embedded in the telecommunication networks are open-loop,

i e. they are based on off-line"reasonable"estimation of network variables (e g. offered traffic), rather than on real-time measurements of such variables.

This limitation is becoming harder and harder, since the telecommunication network behaviours, due to the large variety of supported services and the rapid evolution of the service characteristics, are becoming more and more unpredictable.

This claims for an evolution towards closed-loop algorithms and procedures which are able to properly exploit appropriate real-time network measurements.

In this respect the current technology developments, which assure cheap and powerful sensing capabilities, favours this kind of evolution.

and hence embedding technology-dependent algorithms and procedures, as well as from the large variety of heterogeneous Actors who are playing in the ICT arena.

this framework, on the one hand, is expected to embed algorithms and procedures which, leaving out of consideration the specificity of the various networks,

The concept behind the proposed Future Internet architecture, which aims at overcoming the three above-mentioned limitations,

the proposed architecture is based on a so-called"Cognitive Future Internet Framework"(in the following, for the sake of brevity, simply referred to as"Cognitive Framework")adopting a modular design based on middleware"enablers".

"The enablers can be grouped into two categories: the Semantic Virtualization Enablers and the Cognitive Enablers.

The Cognitive Enablers represent the core of the Cognitive Framework and are in charge of providing the Future Internet control and management functionalities.

They interact with Actors, Resources and Applications through Semantic Virtualization Enablers. The Semantic Virtualization Enablers are in charge of virtualizing the heterogeneous Actors,

Resources and Applications by describing them by means of properly selected, dynamic, homogeneous, context-aware and semantic aggregated metadata.

The Cognitive Enablers consist of a set of modular, technology-independent, interoperating enablers which, on the basis of the aggregated metadata provided by the Semantic Virtualization Enablers,

take consistent control and management decisions concerning the best way to exploit and configure the available Resources

in order to efficiently and flexibly satisfy Application requirements and, consequently, the Actors'needs. For instance, the Cognitive Enablers can reserve network resources,

The control and management decisions taken by the Cognitive Enablers are handled by the Semantic Virtualization Enablers,

Cognitive Future Internet Framework Actors Users Network Providers Prosumer Developers Content Providers Service Providers Applications Semantic Virtualization Enablers Cognitive Enablers Identity

Z Resources Services Networks Contents Devices Cloud storage Terminals Computational Fig. 1. Proposed Cognitive Future Internet Framework conceptual architecture A Cognitive Future

Internet Architecture 95 Note that, thanks to the aggregated semantic metadata provided by the Semantic Virtualization Enablers,

the control and management functionalities included in the Cognitive Enablers have a technology-neutral, multi-layer, multi-network vision of the surrounding Actors, Resources and Applications.

which serve as Cognitive Enabler input, coupled with a proper design of Cognitive Enabler algorithms (e g.,

, multiobjective advanced control and optimization algorithms), lead to cross-layer and cross-network optimization. The Cognitive Framework can exploit one or more of the Cognitive Enablers in a dynamic fashion:

Mobile Terminals, Base Stations, Backhaul network entities, Core network entities. The selection and the mapping of the Cognitive Framework functionalities in the network entities is a critical task

In particular, the concentration of control functionalities in a single framework allows the adoption of algorithms

and more natural. 96 M. Castrucci et al. 3 Cognitive Future Internet Framework Architecture The Cognitive Framework introduced in the previous section consists of a conceptual framework that can be deployed as a distributed functional framework.

It can be realized through the implementation of appropriate Cognitive Middleware-based Agents (in the following referred to as Cognitive Managers)

Core Network entities. There not exist a unique mapping between the proposed conceptual framework over an existing telecommunication network.

However we proposed a proof-of-concept concrete scenario in section 4, where the conceptual framework has been deployed in a real home area network test case.

Indeed the software nature of the Cognitive Manager allows a transparent integration in the network nodes.

that network node is enhanced with the Future Internet functionalities and become part of the Future Internet assets.

Fig. 2 outlines the high-level architecture of a generic Cognitive Manager showing its interfacing with Resources, Actors and Applications.

Fig. 2 highlights that a Cognitive Manager will encompass five high-level modular functionalities, namely the Sensing, Metadata Handling, Elaboration, Actuation and API (Application Protocol Interface) functionalities.

The Sensing, Actuation and API functionalities are embedded in the equipment which interfaces the Cognitive Manager with the Resources (Resource Interface), with the Actors (Actor Interface) and with the Applications (Application Interface;

metadata Enriched data/servicse/contents Monitored Actor related information Aggregated metadata (present context) Exchanged metadata TO/FROM OTHER PEER COGNITIVE MANAGERS Applications API functionalities Application

interface Application protocol Fig. 2. Cognitive Manager architecture A Cognitive Future Internet Architecture 97 The Metadata Handling functionalities are embedded in the so-called Metadata Handling module,

The Metadata Handling and the Elaboration functionalities (and in particular, the Cognitive Enablers which are the core of the proposed architecture) are independent of the peculiarities of the surrounding Resources, Actors and Applications.

With reference to Fig. 2, the Sensing, Metadata Handling, Actuation and API functionalities are embedded in the Semantic Virtualization Enablers,

and of Resource related information (Sensing functionalities embedded in the Resource Interface this monitoring has to take place according to transparent techniques,(ii) the formal description of the above-mentioned heterogeneous parameters/data/services/contents in homogeneous

metadata according to proper ontology based languages (such as OWL Web Ontology Language). Metadata Handling functionalities are in charge of the storing,

(ii) providing enriched data/services/contents to the Actors. In addition, these enablers control the sensing, metadata handling, actuation and API functionalities (these control actions,

for clarity reasons, are represented not in Fig. 2). Actuation functionalities are in charge of (i) actuation of the Cognitive Enabler control decisions over the Resources (Enforcement functionalities embedded in the Resource Interface;

(ii) provisioning to the appropriate Actors the enriched data/contents/services produced by the Cognitive Enablers (Provisioning functionalities embedded in the Actor Interface;

API functionalities are in charge of interfacing the protocols of the Applications managed by the Actors with the Cognitive Enablers.

The concentration of the control functionalities in such Cognitive Enablers allows the adoption of multi-object algorithms and procedures

allow the adoption of innovative and abstract closed-loop methodologies (e g. constrained optimization, data mining, adaptive control, robust control, game theory, operation research, etc.)

for the algorithms and rules embedded in the Cognitive Enablers, which are expected to remarkably improve efficiency.

5) The transparency and the middleware (firmware based) nature of the proposed Cognitive Manger architecture makes relatively easy its embedding in any fixed/mobile network entity (e g.

Mobile Terminals, Base Station, Backhaul network entities, Core network entities: the most appropriate network entities for hosting the Cognitive Managers have to be selected environment by environment.

Moreover, the Cognitive Managers functionalities (and, in particular, the Cognitive Enabler software) can be added/upgraded/deleted through remote (wired and/or wireless) control.

6) The modularity of the Cognitive Manager functionalities allows their ranging from very simple SW/HW/computing implementations,

even specialized on a single-layer/single-network specific monitoring/elaboration/actuation task, to A Cognitive Future Internet Architecture 99 complex multi-layer/multi-network/multi

trading-off the advantages achieved in terms of efficiency with the entailed additional SW/HW/computation complexity.

Internet of things. 8) The above-mentioned flexibility issues favours a smooth migration towards the proposed approach. As a matter of fact, it is expected that Cognitive Manager functionalities will be inserted gradually starting from the most critical network nodes,

we propose to achieve Future Internet revolution through a smooth evolution. In this evolution, Cognitive Managers provided with properly selected functionalities are embedded gradually in properly selected network entities,

which are worthwhile with respect to the increased SW/HW/computing complexity. The following section shows an example of application of the above-mentioned concepts.

where connectivity among devices is provided using heterogeneous wireless (e g.,, Wifi, UWB) and wired (e g.,, Ethernet, PLC) communication technologies.

For 100 M. Castrucci et al. testing purposes only a simplified version of the Cognitive Manager has been implemented in each node of the network,

which includes the following functionalities: the Service and Content adapter: a Qos adapter module has been implemented,

and stored in proper database, ready to be used by elaboration functionalities; a Cognitive connectivity enabler:

it has been implemented to perform technology independent resource management algorithms (e g.,, layer 2 path selection), in order to guarantee that flow's Qos requirements are satisfied during the transmission of its packets over the network.

In particular, a Connection Admission Control algorithm, a Path selection algorithm and a Load Balancing algorithm has been considered in our tests.

The framework has been implemented as a Linux Kernel Module and it has been installed in test-bed machines and in a legacy router1 for performance evaluation.

and two IEEE 802. 3u links at 100 Mbit/s. Fig. 3. Test scenario 1 We have modified the firmware of a Netgear router (Gigabit Open source Router with Wireless

-N and USB port; 453 MHZ Broadcom Processor with 8 MB Flash memory and 64 MB RAM;

a WAN port and four LAN up to 1 Gigabit/s) and cross-compiled the code,

to run the framework on the Router. A Cognitive Future Internet Architecture 101 To test the technology handover performances a FTP download session (file size 175 MB) has been conducted on the Ethernet link.

After approximately 10s, one extremity of the Ethernet cable has been disconnected physically from its socket and the flow has been redirected automatically onto the wireless link thanks a context-aware decision taken by the Cognitive connectivity enabler.

Switching on the Wi-fi link causes more TCP retransmissions and an increased transfer time. This is natural

since Ethernet and Wi-fi have different throughputs. Without the cognitive framework, it is evident that the FTP session would not be terminated at all.

As shown in Fig. 4, the experimented handover time is around 240 ms, during which no packet is received.

Fig. 4. Technology handover 5 Conclusions This paper proposes a novel reference architecture for the Future Internet,

with the aim to provide a solution to overcome current Internet limitations. The proposed architecture is based on Cognitive Modules

thus allowing a smooth migration towards the Future Internet and, at the same time, allowing to implement only the needed functionalities in a give scenario.

Interoperation among heterogeneous entities is achieved by means of their virtualization, obtained thanks to the introduction of Semantic Virtualization Enablers.

At the same time, the Cognitive Enablers, which are the core of the Cognitive Managers, can potentially benefit from information coming from all layers of all networks

and can take consistent and coordinated context-aware decisions impacting on all layers. Clearly which Cognitive Enabler have to be activated,

the algorithms the Cognitive Enabler will be based on, have all to be selected carefully case by case;

The Internet is broken, Technology Review, December 2005-January 2006 (2006), http://www. technologyreview. com/article/16356/2.

Vint Cerf on the Future of the Internet. The Internet Today, The Singularity University (2009), http://www. datacenterknowledge. com/archives/2009/10/12/vint-cerf-on-the-future-of-the-internet/4. National Science Foundation:

Networking Technology and Systems, Nets (2008), http://www. nsf. gov/pubs/2008/nsf08524/nsf08524. htm 5. National Science Foundation:

National Future Internet Initiatives-GRIF (France), http://www. francenumerique2012. fr/(2010) 10. AETIC: Internet del Futuro, http://www. idi. aetic. es/esinternet/(2008) 11.

ICT FP7 Research: Future Internet Research & Experimentation (FIRE), http://cordis. europa. eu/fp7/ict/fire/(2010) Title Model Ontology for Future Internet Networks Joao

Henrique de Souza Pereira1, Flavio de Oliveira Silva1, Edmo Lopes Filho2, Sergio Takeo Kofuji1, and Pedro Frosi Rosa3 1 University of Sao paulo, Brazil joaohs@usp. br, flavio@pad. lsi. usp. br, kofuji@pad. lsi. usp. br

2 Algar Telecom, Brazil edmo@algartelecom. com. br 3 Federal University of Uberlandia, Brazil pedro@facom. ufu. br Abstract.

The currently Internet foundation is characterized on the interconnection of end-hosts exchanging information through its network interfaces usually identified by IP ADDRESSES.

An Internet of active social, mobile and voracious content producers and consumers. Considering the limitations of the current Internet architecture, the envisaged scenarios and work efforts for Future Internet,

this paper presents a contribution for the interaction between entities through the formalization of the Entity Title Model.

Entity, Future Internet, Ontology, Title Model Introduction The Internet of today has difficulties to support the increasing demand for resources

The commercial usage of Internet and IP networks was a considerable obstacle to the improvements in the intermediate layers in this architecture.

The challenges to Future Internet Networks are the primary motivation to this paper and the cooperation in the evolution of computer networks

using the OWL (Web Ontology Language), to collaborate with one integrated reference model for the Future Internet,

including others projects efforts. This paper is organized as follows: Section 1 presents works in the area of Future Internet and ontology in computer systems.

Section 2 describes the concepts of the Entity Title Model and the ontology at network layers.

) Future Internet Assembly, LNCS 6656, pp. 103 114,2011. c The Author (s). This article is published with open access at Springerlink. com. 104 J. H. de

Souza Pereira et al. 1 Future Internet Works A Future Internet full of services requirements demands networks where the necessary resources to service delivery are orchestrated

and projects for the Future Internet and some of these are being discussed in collaboration groups like FIA,

several research groups are working towards a Future Internet reference architecture and the Title Model ontology is a contribution to this area.

the concept of addressing by use of a Title is suitable for real world Internet and its sensor networks.

and subscribe view proposed by PSIRP 30 and used in conjunction with its proposed patterns providing new important inputs to the content-centric view of Future Internet. 1. 1 Some other Future Internet

which seek alternatives to contribute to the evolution of computer networks. In the proposed implementation of LISP there is low impact on existing infrastructure of the Internet

since it can use the structure of IP and TCP, with the separation of Internet addresses into Endpoint Identifiers (EID) and Routing Locators (RLOC) 9. In the area of next generation Internet there is also the works of Landmark developed by Tsuchiya,

that proposed hierarchical routing in large networks and Krioukov work on compact routing for the Internet.

Pasquini proposes changes in the use of Landmark with Rofl (Routing on Flat Labels), and flat routing in binary identity space.

He also proposes the use of domain identifiers for a next-generation Internet architecture 21 22.

and VRR (Virtual Ring Routing) 7. In the area of mobility on a next-generation Internet Wong proposes solutions that include support for multi-homing 36.

there are also proposals Title Model Ontology for Future Internet Networks 105 by Ford, who specifies the UIP/UIA (Unmanaged Internet Protocol) and UIA (Unmanaged Internet Architecture) 12.

Related to ontology, there are extensive studies in philosophy, whose concept of this term is assigned to Aristotle,

However, the name ontology was used first only in the seventeenth century by Johannes Clauberg 2. In the area of technology its initial use was performed by Mealy in 1967 20 and expanded especially in areas of artificial intelligence, database

information systems, software engineering and semantic web. In the technology area one of the most commonly used definitions is from Tom Gruber,

and Semantic web languages (RDF, RDFS, DAML+OIL, OWL SPARQL, GRDDL, RDFA, SHOE AND SKOS), among others 13.

without extending to the middle and lower layers of computer networks. In this research area, this paper aims to contribute to advancing the use of ontology to the intermediate layers as a collaborative proposal for the Future Internet. 2 Ontology at Network Layers Ontologies can use layer model or distinct architectures,

however, in general, they remain restricted to the application layer. For example, the architecture of the Web Ontology Language defined by W3c,

presented in Fig. 1 extracted from 17, is confined in the application layer of the TCP IP architecture.

Fig. 1. Architecture of Web Ontology Language 17.106 J. H. de Souza Pereira et al. In the use of TCP IP, there are limitations concerning the application layer informing its needs to the transport layer.

application, content, host, user, cloud computing and sensor networks. The notion of entity in the Title Model differs from the notion of resources in some relevant literature,

Also can be created other kinds of classification, such as hardware, software and network, among others. Some one of them (not all) can be used as resources in others relevant literature.

and supported by computer networks. For example, in this taxonomy the class layer is a subclass of Thing

Title Model Ontology for Future Internet Networks 107 Title: It is the only designation to ensure an unambiguous identification.

specified in the ISO-9545/X. 207 recommendation, be extended to the other communication entities of the computer networks.

with the purpose of improving the addressing of internet architecture by horizontal addressing and facilitate communication among the entities and with the other layers 24.

It is a tangible material in a computer network, such as: cables, connectors, general optical distributor, antenna, base station and air interface.

to transfer data from a file, or content of email/instant message, it is necessary to have delivery guarantee in communication.

On the other hand, for an audio or video communication in real time, it will not necessarily be important the delivery guarantee,

Possibility of having neighborhoods regardless of physical or logical location of entities in computer networks, without the need of reserved bandwidth, networks segmentation, specific physical connections or virtual private network.

and translate them into functionality in computer networks. Link Layer: This is the layer that has the responsibility to establish the link between two

or more entities and ensure that data exchange occurs at the link level and takes place according to the understanding made by the service layer.

This domain has worldwide coverage and hierarchical scalability formed by elements of local communication, masters and slaves, similar to DNS (Domain name System).

as showed in Fig. 2. 2. 2 Cross Layer Ontology for Future Internet Networks For intermediate semantic layer,

considering others works and projects efforts for Future Internet, as 4ward, Content-Centric, User-Centric, Service-Centric and Autoi Title Model Ontology for Future Internet Networks 109 Source Service Content User DTS

NE1 NE2 NE...NE3 Destination Service Content User Network Elements (NE) Fig. 2. Entities Communication Orchestrated by the DTS. 4 28.

as well as the semantic approaching cross layers for the Future Internet. The Horizontal Addressing by Entity Title has related limitations with the communications needs formalization and standardization,

and also has limitations with the collaboration with others Future Internet projects efforts. The reason is

The benefits for the use of the propositional logic for network formalization is the implementation facility in software and hardware.

in a collaborative effort to others Future Internet works, the Entity Title Model has better contributions by the use of a more expressive and standardized representation language.

For the communication between the layers running in a Distributed Operating system, without the traditional sockets used in TCP IP,

The following OWL sample code shows one use case example for distributed programming, where the application entity with title Master-USP-1 sends its needs to the Service Layer.

Payload Size Control equal to 84 Bytes; and; Delivery Guarantee request. In this context, this need is informed, to the Service Layer,

by the direct use of the Raw Socket to communicate with the Distributed Operating system, without the use of IP, TCP, UDP and SCTP.<

Namedindividual"/>Application title>Master usp 1</Application title><Slave title>Slave usp a</Slave title><Payload size control>84 Bytes</Packet size control><Deliveryguarantee rdf: datatype="&xsd;

comment>Example of the Entity Title Model to support distributed programming needs.</</rdfs: comment><Has need rdf:

Thing>By this semantic information, the Service and Data link layers can support the distributed programming communication using different approaches,

but some of Title Model Ontology for Future Internet Networks 111 them, as Rofl and LISP, should change their structure to semantically support the entities needs

for example, 4ward, Autoi OSKMV planes (Orchestration, Service Enablers, Knowledge management and Virtualisation planes) and the Content-Centric can use this model collaboratively.

In this perspective, the Entity Title Model and its ontology can contribute to converge some Future Internet projects,

In this example for the contribution with the Content, Service and User Centric works, in the Title Model it is possible the unification of the different entities address in the future Internet.

this work aims to contribute with the discussions for a collaborative reference model in the future Internet,

For the service layer to support semantically the entities needs this work uses the Web Ontology Language,

the mobility on the Internet becomes natural, since there is no longer the hierarchy of segments of the network

and sub network that occurs in the IP ADDRESS with the use of masks. By this

In this scenario, this work contributes to the use of ontology in the middle layers of the Internet, with the proposal of semantic formalization, in computer networks, for the Entity Title Model.

This is a possible contribution to the Future Internet efforts and projects like Autoi, Content-Centric, User-Centric

and others, Future Internet efforts. As future work there will be continued the development of this ontology and its collaborative perspective with others Future Internet efforts and projects.

It is suggested to extend discussions and studies concerning the unique identification of the entities and the formalization of security mechanisms for the Entity Title Model.

It is suggested also the continuity of studies and discussions on the use of semantic representation languages in place of protocols in the lower and middle layers of computer networks

Title Model Ontology for Future Internet Networks 113 Open Access. This article is distributed under the terms of the Creative Commons Attribution Noncommercial License

Autonomic Internet Project. European union IST 7th Framework Programme (2011) 5 Bai oco, G.,Costa, A.,Calvi, C.,Garcia, A.:

International Conference on Ultra Modern Telecommunications, IEEE Xplore, Print ISBN: 978-1-4244-3942-3 (2009) 7 Caesar, M.:

Future Internet Design Program. National Science Foundation, http://www. nets-find. net (2011) 11 FIRE:

Future Internet Research and Experimentation (2009) 12 Ford, B.:UIA: A Global Connectivity Architecture for Mobile Personal Devices.

International Journal of Human and Computer Studies, 43 (5 6): 907 928 (1995) 16 ITU-T:

Information technology-Open Systems Interconnection-Application Layer Structure. Recommendation X. 207-ISO/IEC 9545: 1993 (1993) 17 Lacy, L.:

Representing Information Using the Web Ontology Language. Trafford (2005) 18 Lesniewski, S.:Comptes rendus des s'eances de la Soci'et'e des Sciences et des Lettres de Varsovie. pp. 111 132

8th International Information and Telecommunication Technologies Symposium (2009) 114 J. H. de Souza Pereira et al. 20 Mealy, G.:

Another look at data. In: Proceedings of the Fall Joint Computer Conference. AFIPS November 14-16, Volume 31, pp. 525 534.

Thompson Books, Washington and Academic Press, London (1967) 21 Pasquini, R.,Paula, L.,Verdi, F.,Magalh aes, M.:

Domain Identifiers in a Next Generation Internet Architecture. In: IEEE Wireless communications and Networking Conference-WCNC (2009) 22 Pasquini, R.,Verdi, F.,Magalh aes, M.:

Towards a Landmark-based Flat Routing. In: 27th Brazilian Symposium on Computer networks and Distributed systems-SBRC (2009) 23 Pereira, J. H. S.,Kofuji, S. T.,Rosa, P. F.:

Distributed systems Ontology. In: IEEE/IFIP New Technologies, Mobility and Security Conference (2009) 24 Pereira, J. H. S.,Kofuji, S. T.,Rosa, P. F.:

Horizontal Address Ontology in Internet Architecture. In: IEEE/IFIP New Technologies, Mobility and Security Conference (2009) 25 Pereira, J.,Sato, L.,Rosa, P.,Kofuji, S.:

Network Headers Optimization for Distributed Programming. In: 9th International Information and Telecommunication Technologies Symposium (2010) 26 Pereira, J. H. S.,Kofuji, S. T.,Rosa, P. F.:

Horizontal Addressing by Title in a Next Generation Internet. In: IEEE International Conference on Networking and Services p. 7 (2010) 27 Prud encio, A.,Willrich, R.,Diaz, M.,Tazi, S.:

Netqosont: Uma Ontologia para a Especifica¸c ao Sem antica de Qos em Redes de Computadores.

In: 14oworkshop de Ger encia e Opera¸c ao de Redes e Servi¸cos-WGRS-SBRC (2009) 28 Rubio-Loyola, J

Autonomic Internet Framework Deliverable D6. 3. Final Results of the Autonomici Approach. Autoi Project (2010) 29 Souza, J.:

Publish Subscribe Internet Routing Paradigm-PSIRP. Final Updated Architecture, Deliverable D2. 5 (2010) 31 Tselentis, G.,et al.:

Towards the Future Internet-A European Research Perspective. IOS Press, Amsterdam (2009) 32 Tselentis, G.,et al.:

Towards the Future Internet-Emerging Trends from European Research. IOS Press, Amsterdam (2010) 33 Tsiatsis, V.,Gluhak, A.,Bauge, T.,Montagut, F.,Bernat, J.,Bauer, M.,Villalonga, C.,Barnaghi, P

.,Krco, S.:The SENSEI architecture-Enabling the Real world Internet. In: Towards the Future Internet, pp. 247 256.

IOS Press, Amsterdam (2010) 34 Vissers, C.,Logrippo, L.:The Importance of the Service Concept in the Design of Data communications Protocols.

In: Proceedings of the IFIP WG6 1, 3 (1986) 35 Wang, J.,Guo, M.,Camargo, J.:

An Ontological Approach to Computer system Security. Information security Journal: A Global Perspective (2010) 36 Wong, W.:

et al.:An Architecture for Mobility Support in a Next Generation Internet. In: The 22nd IEEE International Conference on Advanced Information, Networking and Applications-AINA (2008) Part II:

Future Internet Foundations: Socioeconomic Issues Part II: Future Internet Foundations: Socioeconomic Issues 117 Introduction Information and Communication Technologies (ICT) provide in recent years solutions to the sustainability challenge by, e g.,

, measuring impacts and benefits of economic activity via integrated environmental monitoring and modeling, by managing consequences,

and by enabling novel low-impact economic activities, such as virtual industries or digital assets. In turn, ICT enables novel systems in terms of technologies

and applications encouraging and generating socioeconomic values. Additionally, these models address in many cases free-market forces,

underpinning the technology of the Internet, is particularly responsible for this accelerating trend. Particularly, controlling and monetizing the evolution of the Internet

and its vast application range is seen as a critical goal for most economic regions. Therefore, socioeconomic aspects determine a highly important set of influencing factors,

While pure economic research as well as pure social research has been undertaken for decades, the combination of the two and its application to the new Internet the one

which is rooted in the commercialization of the native research Internet of the early 90's becomes an important element in investigating,

As collected by the FISE (Future Internet Socioeconomics) working group within the FIA on its wiki, the following general aspects of socioeconomics,

1) The study of the relationship between any sort of economic activity (here networking in the areas of Internet-based and telecommunications-based communications for a variety of lower-level network/telecommunication as well as application-based services) and the social life of user (here,

2) Markets of Internet service providers (ISP) and Telecommunication Providers;(3) ISPS peering agreements and/or transit contracts;(

it clearly outlines that a combination of social and economic viewpoints on pure Internet-based networking is essential.

Thus, the full understanding and modeling of these socioeconomic impacts on Internet communications particularly and the Internet architecture generally challenges networking research and development today.

Future Internet Foundations: Socioeconomic Issues effective approaches. Furthermore, the users'perspectives need to be taken into close consideration,

while being at the same time in contrast to simplicity and easeof-operations of a variety of Internet-based services.

and transmission domain of the Internet had been taken as one starting point of socioeconomic research for this FIA book.

Thus, the content of these chapters on socioeconomics of the Future Internet contains three views

Due to the fact that overlay applications as of today still generate large volumes of data

To ensure a mutually beneficial situation for all stakeholders in a Future Internet scenario, the Triplewin investigations determine the key goal of Economic Traffic Management (ETM) mechanisms developed.

The second chapter by E. Eardly et al. was submitted with the title Deployment and Adoption of Future Internet Protocols.

Based on the assumption that many welldesigned protocols designed for the Future Internet will fail as it happened for the traditional Internet,

as it happens currently for the Future Internet, can get adopted. Finally, the third chapter by C. Kalgoris et al. is on An Approach to Investigating Socioeconomic Tussles Arising from Building the Future Internet.

Based on the assumption that the Internet has evolved into a worldwide social and economic platform with a variety of stakeholders involved,

the key motivations of each of them and their behavior has changed over the recent past dramatically.

Future Internet Foundations: Socioeconomic Issues 119 investigate, classify, and develop an analysis framework for such tussles in the socioeconomic domain of Internet stakeholders.

In turn, the chapter outlines a new methodology, with which tussles are analyzed. Although a survey reveals that many tussles are known,

) Future Internet Assembly, LNCS 6656, pp. 121 131,2011. The Author (s). This article is published with open access at Springerlink. com. Assessment of Economic Management of Overlay Traffic:

Overlay applications generate huge amounts of traffic in the Internet, which determines a problem for Internet service providers,

in order to deal with the overlay traffic in a way that is mutually beneficial for all stakeholders of the Future Internet.

overlays. 1 Introduction Applications such as peer-to-peer (P2p) file sharing and video-streaming generate huge volumes of traffic in the Internet due to their high popularity and large size of the files exchanged.

Thus, besides providing effective solutions for Internet at present ETM is deemed as applicable to the Future Internet, both conceptually and concerning specific ideas and mechanisms.

An example is locality promotion based on BGP routing data. Insertion of Additional Locality-Promoting Peers/Resources involves (a) the insertion of ISP-owned Peers (Iops) in the overlay

For file sharing P2p applications the most important perceivable parameter is download time (or download speed.

if this improves or, at least, preserves their download time. Ideally, this should be guaranteed on a per individual user basis. However,

Another dimension of a service provider's win is decreased a traffic volume from its own content servers and reduced load of the servers,

and reflecting a part of the real Internet topology, with a subset of ASES and inter-domain connections;

Peers and the ETMS servers, providing rating information, are located in these stub-ASES, which are interconnected via a hub-AS containing the initial seed.

15 20 9 10 11 12 13 14 15 AS ID Download Times (min) Ref BGPLOC Fig. 1. Mean Inter-AS Bandwidth

Fig. 2. Download Times On the other hand, the situation is not as simple when considering end users, cf.

but an increase of the outgoing traffic due to the data exchange also with remote peers;

Fig. 3. Mean Inter-AS Bandwidth Fig. 4. Download Times Swarm Selection. In the underlay considered the same setup was applied;

Fig. 5. Mean Inter-AS Bandwidth Fig. 6. Download Times 128 I. Papafili et al. Table 1. Evaluation Scenarios for the Swarm Selection Scenario A b c Modified parameters File Size:

200.0 s Fig. 5 and Fig. 6 present results for the inter-domain traffic and for the peers'download times, respectively.

The main evaluation goal is to show by simulations that the HAP ETM mechanism allows for decreasing download times of peers that want to become HAPS (due to the extra download bandwidth offered to them.

the mean download time decreases significantly (see difference between‘No SIS'and‘0 HAPS');'this is due to the fact that peers in AS1 share their resources among themselves, in contrast to the rest of peers,

Afterwards, the download time can be reduced further by increasing the number of active HAPS. This phenomenon can be justified,

In the former case, the mean download time is reduced more than in the latter. This is due to the fact that the injection of,

when less peers are present in the AS, hence the difference in download time. The results in Fig. 7 clearly show that end-users benefit from the introduction of HAP ETM mechanism.

and have gain additional download bandwidth, but also, with their extra upload bandwidth HAPS lead to the significant reduction of the average download time too.

Fig. 7. Mean Download Times for Peers in AS1 With respect to the Number of HAPS When an HAP is implemented along with locality-awareness mechanisms,

the operator benefits from reduced inter-domain traffic 17. It allows for reducing costs for ISPS

Implementation-wise for an operational prototype, the Admin component of the Smoothit Information Service (SIS) has been designed as a Web-based tool for the ISP to administrate

Finally, the extension and application of the methodology for other traffic types (not only P2p) generated according to trends in the future Internet is an interesting and promising direction for future research.

IEEE International Conference on Peer-to-peer Computing P2p 2010, Delft, The netherlands (August 2010) 7. Bindal, R.,Cao, P.,Chan, W.,Medval, J

26th IEEE International Conference on Distributed computing Systems, Montreal, Canada (June 2006) Assessment of Economic Management of Overlay Traffic:

9th International Conference on Peer-to-peer Computing (P2p'09), Seattle, USA (September 2009) 9. The Smoothit Project:

19th IEEE International Conference on Computer Communications and Networks (ICCCN 2010), Zürich, Switzerland (August 2010) 11.

Characterization of Bittorrent Swarms and their Distribution in the Internet, to appear in the Computer networks (2011) 13.

) Future Internet Assembly, LNCS 6656, pp. 133 144,2011. The Author (s). This article is published with open access at Springerlink. com. Deployment and Adoption of Future Internet Protocols Philip Eardley1, Michalis Kanakakis2, Alexandros Kostopoulos2, Tapio Levä3, Ken Richardson4,

and Henna Warma3 1 BT Innovate & Design, UK philip. eardley@bt. com 2 Athens University of Economics and Business, Greece {kanakakis, alexkosto}@ aueb. gr

Many, if not most, well-designed Future Internet protocols fail, and some badly-designed protocols are very successful.

Careful consideration of such issues can increase the chances that a future Internet protocol is adopted widely.

a good example being GSM but there are many more examples of protocols that are designed well technically but where deployment has failed

Several attempts have been made at studying the adoption of consumer products 1 and new Internet protocols,

The adoption of Internet protocols is tricky because the Internet is a complex system with diverse end-systems, routers and other network elements, not all of

whose aspects are under the direct control of the respective end users or service providers. In this Chapter we propose a new framework for a successful adoption process (Section 2),

and Adoption of Future Internet Protocols We propose a new framework (Figure 1) for a successful adoption process, with several key features:

Network effect Testing Testing Testing Fig. 1. An adoption framework Deployment and Adoption of Future Internet Protocols 135 A version of the framework has been applied in two papers,

8 and 9. The framework is intended to be generally applicable to Internet protocols. The first key question is:

browsers and the underlying http/html protocols give a significant benefit to both the end users (a nice user interface for easy access to the web)

As a counterexample, IPV6 deployment has a cost to the end host to support the dual stack,

As a counterexample, IPV6 requires at least both ends and preferably the network to change. Combining these two key factors leads to the idea of an incremental process, where the aim at each step is to bring a net benefit for the party (s) migrating.

since the software has already been developed for the initial scenario and it is simply a matter of deploying

The current Internet's routing system only exposes a single path between a source-address pair

for example mobile devices have multiple interfaces. MPTCP supports the use of multiple paths between source and destination.

when both endpoints understand the necessary Deployment and Adoption of Future Internet Protocols 137 extensions to support MPTCP.

again, this helps persuade the IETF that it is safe to deploy on the internet.

it just uses TCP's API so it looks the same to applications. This, plus the following two bullets, help MPTCP be incrementally deployable.

It is designed to be middlebox-friendly (be it a NAT, firewall, proxy or whatever), in order to increase the chances that MPTCP works

So we are now working on a NAT survey to probe random paths across the Internet to test how operational NATS impact MPTCP's signalling messages 21.

our current belief is that a data centre is the most promising initial scenario (Figure 2). Within a data centre,

one issue today is how to choose what path to use between two servers amongst the many possibilities-MPTCP naturally spreads traffic over the available paths.

Simulations show there are significant gains in typical data centre topologies 25, perhaps increasing the throughput from 40%to 80%of the theoretical maximum.

However, the protocol implementation should not impact hardware offloading of segmentation and check-summing. One reason that MPTCP uses TCP-Options for signalling (rather than the payload) is that it should simplify offloading by network cards that support MPTCP,

due to the separate handling of MPTCP's signalling and data. Incremental: the story is good,

as only one stakeholder is involved viz the data centre operator. Fig. 2. Potential MPTCP deployment scenario, in a data centre.

In this example, traffic between the two servers (at the bottom) travels over two paths through the switching fabric of the data centre (there are four possible paths.

Another potential initial scenario would be a mobile user using MPTCP over multiple interfaces. The scenario reveals a potential distinction between deployment

(which involves the OS vendor updating their stack) and adoption (which means that MPTCP is actually being used

and deployment is decided mainly by the OS (Operating system) vendor and not the end user.)Therefore we believe that a more promising initial scenario is an end user that accesses content, via wireless LAN and 3g, from a provider that controls both end user devices and content servers 26 for example,

Nokia or Apple controls both the device and the content server, Nokia Ovi or Apple App store.

Benefits: MPTCP improves resilience -if one link fails on a multi-homed terminal, the connection still works over the other interface.

Both the devices and servers are under the control of one stakeholder, so the end user‘unconsciously'adopts MPTCP.

However, there may be NATS on the data path, and MPTCP's signalling messages must get through them.

Deployment and Adoption of Future Internet Protocols 139 The wider scenario of widespread deployment and adoption is again worth thinking about this even during the design of the protocol.

For instance, it is necessary to think about the benefits and costs for OS vendors, end users, applications and ISPS (Internet service providers.

For instance, as soon as a major content provider, such as Google, deploys MPTCP perhaps as part of a new application with better Qos-then there is a much stronger incentive for OSS to deploy it as well as the network externality has increased suddenly.

ie the API is unaltered (although there will also be enhanced an API for MPTCP-aware applications).

MPTCP is an extension for end-hosts it doesn't require an upgrade to the routing system;

if both ends of the connection have deployed MPTCP, then it just works (NATS permitting). 4 Congestion Exposure The main intention of Congestion Exposure (Conex) is to make users

In today's internet this information is only visible at the transport layer, and hence not available inside the network without packet sniffing.)

and to decrease the latency of data delivery. The CDN server sends premium packets (perhaps for IPTV) as Conex-Not-Marked or Conex-Re-echo.

Conex traffic is prioritised by the operator (premium service. To a first order of approximation, the only point of contention is the backhaul where the operator already has a traffic management box,

The operator upgrades its traffic management box so that it drops Conex traffic with a lower probability.

Only one party has to upgrade, ie the combined CDN-ISP. The Content providers and consumers don't know about Conex.

Deployment and Adoption of Future Internet Protocols 141 One way this scenario could widen out is that the content provider is informed now about the Conex-Re-echoes

Therefore the ISP needs to upgrade two things. Firstly its traffic management box: it needs to do occasional auditing spot-checks,

and then the host's software would automatically send the user's premium traffic (Voip say) as Conex-enabled.

or end user at a time. 5 Enhancing the Framework One important development in telecoms is virtualisation. Although the basic idea is longstanding,

Roll out of the software should be cheaper, therefore the expected benefits of the deployment can be less.

There is no need to coordinate end users all having to upgrade. Every user can immediately use the new (virtualised) software,

so effectively a large number of users can be enabled simultaneously. These factors reduce the deployment risk,

if there is some problem with the new software. Virtualisation is not suitable for all types of software, for instance new transport layer functionality, such as MPTCP and CONEX,

needs to be on the actual devices. 142 P. Eardley et al. There is an analogy with the digitalisation of content

Virtualisation should similarly lower the cost of distribution in other words, it eases deployment. Another aspect is the interaction of a new protocol with existing protocols.

One set of examples is the various IPV4-IPV6 transition mechanisms that try to release the (currently hidden) benefits of IPV6.

RFC 5218 (2008) Deployment and Adoption of Future Internet Protocols 143 3. Burness, L.,Eardley, P.,Akhtar, N.,Callejo, M. A.,Colas

a key requirement for systems beyond 3g. In: VTC 2005-Spring, IEEE 61st Vehicular Technology Conference (2005) 4. Hovav, A.,Patnayakuni, R.,Schuff, D.:

A model of Internet Standards Adoption: the Case of IPV6. Information systems Journal 14 (3), 265 294 (2004) 5. Katz, M.,Shapiro, C.:

Technology Adoption in the Presence of Network Externalities. Journal of Political Economics 94,822 841 (1986) 6. Joseph, D.,Shetty, N.,Chuang, J.,Stoica, I.:

Modeling the Adoption of New Network architectures. In: International Conference on Emerging Networking Experiments and Technologies (2007) 7. Dovrolis, C.,Streelman, T.:

ACM SIGCOMM Computer Communications Review 40 (2)( 2010) 8. Kostopoulos, A.,Warma, H.,Leva, T.,Heinrich, B.,Ford, A.,Eggert, L.:

NGI 2010-6th Euronf Conference on Next Generation Internet, Paris (2010) 9. Kostopoulos, A.,Richardson, K.,Kanakakis, M.:

IEEE International Conference on Network protocols, ICNP (2002), http://www. ece. gatech. edu/research/GNAN/work/ptcp/ptcp. html 14.

Internet Conference IC (2002) 15. Zhang, M.,Lai, J.,Krishnamurthy, A.,Peterson, L.,Wang, R.:

Adding concurrent data transfer to transport layer, Proquest ETD Collection for FIU, Paper AAI3279221 (2007), http://digitalcommons. fiu. edu/dissertations/AAI3279221 17.

Improved data distribution for multipath TCP communication. In: IEEE GLOBECOM (2005) 19. Kelly, F.,Voice, T.:

Stability of end-to-end algorithms for joint routing and rate control. Computer Communication Review 35,2 (2005) 20.

Key, P.,Massoulie, P.,Towsley, D.:Combined Multipath Routing and Congestion Control: a Robust Internet Architecture, no.

MSR-TR-2005-111 (2005), http://research. microsoft. com/pubs/70208/tr-2005-111. pdf 21.

Honda, M.:Call for contribution to middlebox survey (2010), http://www. ietf. org/mail-archive/web/multipathtcp/current/msg01150. html 22.

Becke, M.,Dreibholz, T.,Iyengar, J.,Natarajan, P.,Tuexen, M.:Load Sharing for the Stream Control Transmission Protocol (SCTP), draft-tuexen-tsvwg-sctp-multipath-00. txt, work in progress (2010) 144 P. Eardley et al. 23.

HTTP Extensions for Simultaneous Download from Multiple Mirrors, draft-ford-http-multi-server, work in progress (2009) 25.

Data center Networking with Multipath TCP. ACM Sigcomm Hotnets (2010) 26. Warma, H.,Levä, T.,Eggert, L.,Hämmäinen, H.,Manner, J.:

Mobile Internet In Stereo: an End-to-end Scenario. In: 3rd Workshop on Economic Traffic Management, ETM (2010) 27.

) Future Internet Assembly, LNCS 6656, pp. 145 159,2011. The Author (s). This article is published with open access at Springerlink. com. An Approach to Investigating Socioeconomic Tussles Arising from Building the Future Internet Costas Kalogiros1, Costas Courcoubetis1, George D

. Stamoulis1, Michael Boniface2, Eric T. Meyer3, Martin Waldburger4, Daniel Field5, and Burkhard Stiller4 1 Athens University of Economics and Business, Greece ckalog@aueb. gr, courcou@aueb. gr, gstamoul@aueb. gr 2 University of Southampton

IT Innovation, United kingdom mjb@it-innovation. soton. ac. uk 3 Oxford Internet Institute, University of Oxford, United kingdom eric. meyer@oii. ox. ac. uk

With the evolution of the Internet from a controlled research network to a worldwide social and economic platform, the initial assumptions regarding stakeholder cooperative behavior are no longer valid.

Future Internet Socioeconomics, Incentives, Design Principles, Tussles, Methodology 1 Introduction The Internet has moved already long since from the original research-driven network of networks into a highly innovative, highly competitive marketplace for applications

Accordingly, different stakeholders in the Internet space have developed a wide range of on-line business models to enable sustainable electronic business.

Furthermore, the Internet is increasingly pervading society 3. Widespread access to the Internet via mobile devices, an ever-growing number of broadband users worldwide,

and trends like the Internet-of-Things or the success of Cloud services, all provide indicators of the high significance of the Internet today.

Hence, social and economic impacts of innovations in the future Internet space can be reasonably ex 146 C. Kalogiros et al. pected to increase in importance.

Thus, since the future Internet can be expected to be characterized by an ever larger socioeconomic impact,

and using the Internet. That is a tussle is a process in which each stakeholder has particular self-interests,

and mobile network convergence 10 constitute only two representative examples for typical tussle spaces. The main argument for focusing on tussles in relation to socioeconomic impact of the future Internet is in the number of observed stakeholders in the current Internet and their interests.

Clark et al. speak of tussles on the Internet as of today. They argue 5 that t here are,

and have been for some time, important and powerful players that make up the Internet milieu with interests directly at odds with each other.

With the ongoing success of the Internet and with the assumption of a future Internet being a competitive marketplace with a growing number of both users and service providers,

tussle analysis becomes an important approach to assess the impact of stakeholder behavior. This paper proposes a generic methodology for identifying

such as the current and future Internet. In order to help an analyst during the tussle identification task

when introducing new information systems into organizations, suggests an iterative approach to studying complex and problematic real-world situations (called systems)

and analyzing socioeconomic tussles in the future Internet. In Section 3 we provide a classification of tussles according to stakeholders'interests into social and economic ones,

and Assessing Tussles The Design for Tussle goal is considered to be a normal evolution of Internet design goals to reflect the changes in Internet usage.

This paradigm shift should be reflected in new attempts for building the Future Internet. However, identifying both existing and future socioeconomic tussles,

Providing a systematic approach for this task has received little attention by Future Internet researchers. Such a methodology should be a step-by-step procedure that can be applied to any Internet functionality

acting as a guide for making sure that all important factors are considered when making technology decisions.

For example the latter could apply this methodology before and after protocol introduction in order to estimate the adoptability and other possible effects, both positive and negative ones, for the Future Internet.

since there is no API (Application programming interface) for ASPS to affect how their traffic will be handled. ASPS and HUS can employ protocol obfuscation techniques

The characteristics of each pattern can be seen in many current and future Internet scenarios. Each pattern looks at relationships between consumers and suppliers and how conflicts of interest can emerge through technical innovations.

For instance, while individual Internet users are typically consumers, when they are creating data that a business would like to sell, with or without their knowledge and consent,

they are providers of the resource in such a scenario. The initial set of tussle patterns is described below.

It is important to note that many innovations in the Internet space have involved repurposing of resources,

and radio frequencies shared between users and wireless devices. In the former case, modern transport control protocols perform congestion control without considering the utility of the sender on instantaneous bandwidth or the number of their active connections.

such as processing and storage capabilities of servers and networking infrastructure. For example, routing table memory of core Internet routers can be considered a public good that retail ISPS have an incentive to over-consume by performing prefix de-aggregation with Border Gateway Protocol (BGP.

Another type of scarce Internet resources is network identifiers, like IPV4 addresses and especially Provider Independent ones that ease net An Approach to Investigating Socioeconomic Tussles 153 work management and avoid ISP lock in.

Sometimes a contention tussle between consumers can have side effects on the owner of the scarce resource,

which is an economic entity and must protect its investments. Examples include the deployment of Deep Packet Inspection techniques by ISPS

when a set of providers collaborate during service provision with strict requirements, like long-distance phone conversations taking place over Internet.

In the example of file sharing applications, if an ISP deployed middle-boxes and performed traffic shaping then it may have negative impact on the services,

and reliability asking for a backup path towards a destination, or prefer 154 C. Kalogiros et al. avoiding specific upstream ISPS.

when multiple candidate servers are available, a consumer may prefer the one offering better Qos,

while a provider selects the server that minimizes its cost; e g.,, this is possible if the provider operates a local DNS service.

Repurposing tussles occur in regards to the privacy of user communication data between users, ISPS, service providers and regulators.

they must be given access to network communication data. Furthermore, ISPS and other companies such as Google and Amazon have increasingly been able to monetize their user transaction data and personal data.

Google is feed able to advertisements based on past searching and browsing habits, and Amazon is able to make recommendations based on viewing and purchasing habits.

These applications of user data as marketing tools are largely unregulated. And in many cases, users have proved willing to give up some of their privacy in exchange for the economic benefit of better deals that can come from targeted advertising.

However, for users who wish to opt out of such systems, the mechanisms for doing so are often less than clear,

Responsibility tussles occur with ISPS that often inhabit a middle ground they are the bodies with direct access to the data

particularly when compared to the speed of change in many technological systems such as the Future Internet.

what technical designs can protect such sites from being attacked by entities inconvenienced or embarrassed by their revelations?

The Internet makes this a particularly contentious issue because with the global nature of the Internet one can't just assume Western values

(as if it were possible even within Europe to agree to what that means). Where does national sovereignty fit into all of this?

The Trilogy project 16 studied extensively the contention tussle among users as well as among an ISP and its customers, due to the aggressive behavior of popular file sharing applications.

and a novel con 156 C. Kalogiros et al. gestion control algorithm that gives the right incentives to users of bandwidth intensive applications.

the attempt to acquire sensitive personal data of end-users by masquerading as a trustworthy entity, as a reverse contention tussle among two website owners (the consumers).

The tussle is being played out in the routing domain: the fraudulent one advertises more specific BGP prefixes

so that ISPS update the entries in their routing tables (the resource) and route end-user requests to the fake website instead of the real one.

The Smoothit project (Simple Economic Management Approaches of Overlay Traffic in Heterogeneous Internet Topologies) studies the control tussle that arises between ISPS

. Since it is based on a lean architecture to operate new services in the future Internet, the discovery of capabilities and the adaptation of many management operations to current working An Approach to Investigating Socioeconomic Tussles 157 conditions of a network are major elements in the new approach.

where embedded capabilities of networking devices and elements see defaulton management functionality, which consist out of autonomous components interacting with each other in the same device and with components in neighboring devices.

Due to economies of scale the thin-client paradigm, where most applications run on a remote server, is considered to achieving energy savings but to the disadvantage of the server provider.

However under some assumptions, Wifi hotspots can consume much less energy than UMTS (Universal mobile telecommunications system) networks.

Thus, responsibility cannot be checked easily. Furthermore, this situation triggers a control tussle between wireless network operators and users of dual-band devices (e g.

Wifi and UMTS) on the technology used to communicate. Next generation networks, where a provider can control

as a result of technological changes and innovations being researched to advance the Future Internet. One challenge for the technologists designing new hardware, software systems,

and platforms, however, is to be aware that technology is not value-free, since it can have several consequences.

To some extent, this message has already been taken on board by many policy makers, computer scientists, and systems designers.

and the Future Internet research community by offering selected services to FP7 projects in Challenge 1. SESERV provides access to socioeconomic experts investigating the relationship between FI technology, society,

and assessing tussles that are present in the Internet, or may arise after a protocol

and WIMAGIC try to design technical solutions that achieve efficient spectrum usage for mobile devices. Following the increasing consensus on benefits of incorporating economic incentive mechanisms in technical solutions, several projects like Trilogy, Smoothit, ETICS,

when making technology decisions and/or the inherent difficulty of addressing socioeconomic issues in the Internet when such challenges still exist in the real world.

Towards a Future Internet: Interrelation between Technological, Social and Economic Trends, Final Report for DG Information Society and Media, European commission DG INFSO, Project SMART 2008/0049 (2010) 3. Blazic

The Future of the Internet: Tussles and Challenges in the Evolution Path as Identified. In:

Defining Tomorrow's Internet. IEEE/ACM Transactions on Networking 13 (3), 462 475 (2005) 6. Courcoubetis, C.,Weber, R.:

New Design Principles for the Internet. In: IEEE International Conference on Communications Workshops, June 2009, pp. 1 5 (2009) 10.

16th Annual Symposium on Theoretical Aspects of Computer science 1999, pp. 404 413 (1999) 13. MOBITHIN project:

Towards the Future Internet-Emerging Trends from European Research, IOS Press, Amsterdam (2010) 16. Trilogy:

Future Internet Foundations: Security and Trust Part III: Future Internet Foundations: Security and Trust 163 Introduction If you are asking for the major guiding principles of Future Internet technology and applications,

the answer is likely to include sharing and collaboration. Cloud computing, for instance, is built on shared resources and computing environments,

offering virtualized environments to individual tenants or groups of tenants, while executing them on shared physical storage and computation resources.

The concept of Platform-as-a-service provides joint development and execution environments for software and services, with common framework features and easy integration of functionality offered by third parties.

The Internet of Services allows the forming of value networks through on-demand service coalitions built upon service offerings of different provenance and ownership.

And, finally, the principle of sharing and collaboration reaches to the applications and business models, ranging from the exchange of data of physical objects for the optimization of business scenarios in, e g.,

, retail, supply chain management or manufacturing, the Internet of things to social networks. While it is evident that sharing

and collaboration brings the Internet, its technologies, applications and users to the next level of evolution,

it also raises security and privacy concerns and introduces additional protection needs. The Future Internet is characterized by deliberate exposure of precious information

and resources on one hand and a number of likely previously unknown interacting entities on the other hand, including service and platform providers as well as service brokers and aggregators.

Valuable and sensitive information, be it business or personal data, should, however, only be exposed to known and trusted entities

and in a controlled way, allowing the owner of the data to decide and control how,

when, and where it is going to be used. These are not new requirements in nature, but the Future Internet adds new dimensions of scale and complexity.

The number of participating and collaborating entities reaches billions when we consider the inclusion of physical objects,

Data travel through a multitude of different domains, contexts and locations while being processed by a large number of entities with different ownership.

and treated according to the data owner's policy, in balance with the processing entities'policies.

infrastructure provider or service broker or any other Future Internet entity, while distribution and exchange of data serve for additional entry points that can potentially be exploited to penetrate a system.

The challenge is to design security and trust solutions that scale to Future Internet complexity and keep the information and resource owner in control, balancing potentially conflicting requirements while still supporting flexibility and adaptation.

Explicit specification of protection needs in terms of declarative policies is key, as well as providing assurance about security properties of exposed services and information. 164 Part III:

Future Internet Foundations: Security and Trust The chapters presented in the Security and Trust section of this volume look at the challenges mentioned above from three different angles.

First, Future Internet principles are supported by revised communication paradigms, which address potential security issues from the beginning,

The chapter, Security Design for an Inter-domain Publish/Subscribe Architecture by K. Visala et al. looks into security implications of a data-centric approach for the Future Internet,

and scoping that ensure the availability of data and maintains their integrity. It is a good example of how clean-slate approaches to the Future Internet can support security needs by design,

rather than provided as an add-on to an existing approach, as has been the case for the current Internet.

The second group of chapters investigates the provision of assurance of the security properties of services and infrastructures in the future Internet.

The provision of evidence and a systematic approach to ensure that best security practices are applied in the design

and operation of Future Internet components are essential to provide the needed level of trustworthiness of these components.

The chapter Engineering Secure Future Internet Services by W. Joosen et al. makes a point for establishing an engineering discipline for secure services,

taking the characteristics of the Future Internet into account. Such a discipline is required to particularly emphasize multilateral security requirements, the composability of secure services,

The authors propose security support in programming and execution environments for services, and suggest using rigorous models through all phases of the SDLC, from requirements engineering to model-based penetration testing.

Their considerations lead to the identification of Future Internet specific security engineering research strands. One of the major ingredients of this program, the provision of security assurance through formal validation of security properties of services, is investigated in detail in the chapter‘Towards Formal Validation of Trust and Security in the Internet of Services by R

. Carbone et al. They introduce a language to specify the security aspects of services and a validation platform based on model-checking.

A number of distinguished features ensure the feasibility of the approach to Future Internet scenarios and the scalability to its complexity:

and trust assurance in the future Internet addressing one of the major obstacles preventing businesses and users to fully exploit the Future Internet opportunities today.

While engineering and validation approaches provide a framework for the secure design of Future Internet artifacts adapted to its characteristics, the third group of Part III:

Future Internet Foundations: Security and Trust 165 chapters looks into specific instances of the information sharing and collaboration principle and introduces novel means to establish their security.

The chapter Trustworthy Clouds underpinning the Future Internet of R. Glott et al. discusses latest trends in cloud computing and related security issues.

The vision of clouds-of-clouds describes collaboration and federation of independent cloud providers to provide seamless access to end users,

and provide an outlook to their mitigation, embedded in a systematic security risk management process. In cloud computing,

but also in most other Future Internet scenarios like the Internet of Services, the need for data exchange leads to sensitive data, e g.,

, personally identifiable information, travelling across a number of processes, components, and domains. All these entities have the means to collect

and exploit these data, posing a challenge to the enforcement of the users'protection needs and privacy regulations.

This is amplified by the dynamic nature of the Future Internet, which does not allow one to predict by whom data will be processed

or stored. To provide transparency and control of data usage the chapter Data Usage Control in the future Internet Cloud proposes a policy-based framework for expressing data handling conditions

and enforcing them. Policies relating events and obligations are coupled with data (sticky policies) and, hence, cannot get lost in transition.

A common policy framework based on tamper-proof event handlers and obligation engines allows for the evaluation of user-defined policies

and their execution, leaving control to the user. With the three groups of chapters, this section of the book provides directions on how security

and collaboration in the future Internet can be mitigated, removing a major hurdle for using its exciting opportunities in sensitive scenarios of both the business and societal worlds.

) Future Internet Assembly, LNCS 6656, pp. 167 176,2011. The Author (s). This article is published with open access at Springerlink. com. Security Design for an Inter-Domain Publish/Subscribe Architecture Kari Visala1, Dmitrij Lagutin1,

and Sasu Tarkoma2 1 Helsinki Institute for Information technology HIIT/Aalto University School of Science and Technology, Espoo, Finland {Kari.

Several new architectures have been proposed recently to replace the Internet Protocol Suite with a data-centric

or publish/subscribe (pub/sub) network layer waist for the Internet. The clean-slate design makes it possible to take into account issues in the current Internet

such as unwanted traffic, from the start. If these new proposals are deployed ever as part of the public Internet as an essential building block of the infrastructure,

they must be able to operate in a hostile environment, where a large number of users are assumed to collude against the network and other users.

In this paper we present a security design through the network stack for a data-centric pub/sub architecture that achieves availability, information integrity,

Future Internet, publish/subscribe networking, network security 1 Introduction Data-centric pub/sub as a communication abstraction 2, 3,

4 reverses the control between the sender and the receiver. Publication in the middle decouples the publisher from the subscriber and there is no direct way of sending a message to a given network,

but our goal is to replace the whole Internet protocol suite with a clean-slate data-centric pub/sub network waist 14.

For example, it must be assumed that the core routers forward packets at line-speeds of tens of Gigabits per second

and minimal in complexity and trust assumptions between stakeholders. 2 Basic Concepts Data-or content-centric networking can be seen as the inversion of control between the sender

the receiver expresses its interest in some identified data that the network then returns when it becomes available taking advantage of multicast

3. We use the term information-centric for this communication pattern to emphasize that the data items can link to other named data

and that the data has structure. An immutable association can be created between a rendezvous identifier (Rid)

and a data value by a publisher and we call this association a publication. At some point in time, a data source may then publish the publication inside a set of scopes that determine the distribution policies such as access control

routing algorithm, reachability, and Qos for the publication and may support transport abstraction specific policies such as replication and persistence for data-centric communication.

The Security Design for an Inter-Domain Publish/Subscribe Architecture 169 scope must be trusted by the communicating nodes to function as promised and much of the security of our architecture is based on this assumption as we explain in 5. Scopes are identified with a special type

of Rid called scope identifier (Sid. Even though the control plane of our architecture, implementing the rendezvous function,

operates solely using data-centric pub/sub model, it can be used to set up communication using any kind of transport abstraction on the data plane fast path,

that is used for the payload communication. The data-centric paradigm is a natural match with the communication of topology information that needs to be distributed typically to multiple parties

and the ubiquitous caching considerably reduces the initial latency for the payload communication as popular operations can be completed locally based on cached data.

Below the control plane t he network is composed of domains, that encapsulate resources such as links,

storage space, processing power in routers, and information. The concept of domain is here very general,

and can refer to abstractions of any granularity, such as software components, individual nodes, or ASES.

the roles for the endpoints are a source and a destination or for data-centric transport:

a data source and a subscriber. The topic is identified with an Rid and is used to match the end nodes in correct interaction instances by the scope.

For example, for data-centric communication, the topic identifies the requested publication. A graphlet defines the network resources used for the payload communication

and L is a variable length label of binary data. Only fixed length hash of the identifier is used in-network

where the data source uses the label as an argument to produce the publication on the fly.

Fig. 1 depicts a simplified example of My movie edit meta-data publication that has Rid (PN

The contents of this publication point to another movie frame data publication indirectly using a so called application level identifier (Aid) of the referred publication.

An FPGA based hardware accelerator has been developed for PLA 24 accelerating cryptographic operations. Security Design for an Inter-Domain Publish/Subscribe Architecture 171 Fig. 1. Publications can refer to other publications persistently using long-term Aids.

where publications are made available are orthogonal to the structure of the data. In Fig. 1, the publication on the left is published inside My home scope that is fully controlled by the local user.

In this example, it is easy to see that the logical structure of the data, e g. the link between the two publications, is orthogonal to the scoping of the data that determines the communication aspects for each publication. 2. 2 Interdomain Structure Each node has an access to a set of network resources In the current Internet,

most policy compliant paths have the so-called valley-free property 16, which means that, on the AS business relationship level,

which implements a data-centric pub/sub primitive as a recursive, hierarchical structure, which first joins node local rendezvous implementations into rendezvous networks (RN)

In another dimension, the rendezvous system is split into common rendezvous core and scope-specific implementations of scope home nodes that implement the functionality for a set of scopes.

and produces an endtoend path between the service container e g. a data source) and the client (e g. a subscriber) and returns the information to the client that can then use this information to join a graphlet (e g. a delivery tree) that can then be used for the fast-path payload communication.

in order to keep the publica Security Design for an Inter-Domain Publish/Subscribe Architecture 173 tion data or pending subscription alive.

This pub/sub primitive is the only functionality implemented by the rendezvous core. We refer to our work in 5 for a detailed description of the rendezvous security mechanisms.

When a cached result cannot be found in the rendezvous core, the subscription reaches the scope,

but this type of applications should be supported by adding a data-centric transport to the data plane as we did in 2. Topology manager (TM) is another function that is implemented by each independently managed domain.

Each scope also publishes a meta-data publication inside itself named (DKX, scope meta-data) describing which transports the scope supports, among others.

It should be noted that the upgraph combination based routing does not require any type of central entity to manage addresses

The upgraph data itself is published by the provider domain of the node. Because many nodes share the same upgraph,

the data-centric rendezvous system caches them orthogonally close to the scope homes that are nodes implementing the scope in question.

If the transport in question is multicast data dissemination then a separate resource allocation protocol could be coupled with the protocol as we did in 2. The client side implementation of the transport would then take the resource description from rendezvous as an input

A data-oriented network architecture DONA 4 replaces a traditional DNS-based namespace with self-certifying flat labels,

which owns the data and L is a label. DONA utilizes an IP header extension mechanism to add a DONA header to the IP header,

Consumers of data send interest packets to the network, and a nodes possessing the data reply with the corresponding data packet.

Since packets are named independently, a separate interest packet must be sent for each required data packet.

In CCN data packets are signed by the original publisher allowing independent verification, however interest packet's are protected not always by signatures.

Security issues of the content-based pub/sub system have been explored in 7. The work proposes secure event types

where the publication's user friendly name is tied to the publisher's cryptographic key. Security Design for an Inter-Domain Publish/Subscribe Architecture 175 5. 1 Security Mechanisms Most of existing network layer security proposals utilize hash chains

Accountable Internet Protocol (AIP) 11 aims to improve security by providing accountability on the network layer.

If the router receives a packet from the unknown EID, the router will send a verification message back

and the node will reply with a message signed by its private key. Since EID is hash of node's public key,

and Future Work In this paper we introduced a data-centric inter-domain pub/sub architecture addressing availability and data integrity.

We used the concept of scope to separate the logical structure of linked data from the orthogonal distribution strategies used to determine how the data is communicated in the network.

Security issues and requirements for Internet-scale publish-subscribe systems. In: HICSS'02, Hawaii, USA (2002) 2. Visala, K.,Lagutin, D.,Tarkoma, S.:

An Inter-Domain Data-Oriented Routing Architecture. In: Rearch'09, Rome, Italy (2009) 3. Jacobson, V.,Smetters, D. K.,Thornton, J. D.,Plass, M.,Briggs, N.,Braynard, R. L.:

A Data-Oriented (and Beyond) Network architecture. In: ACM SIGCOMM 2007, Kyoto, Japan (2007) 176 K. Visala, D. Lagutin,

Defining Tomorrow's Internet. IEEE/ACM Transactions on Networking 13 (3), 462 475 (2005) 7. Pesonen, L. I.,Bacon, J.:

5th international workshop on Software engineering and middleware, pp. 98 105 (2005) 8. Merkle, R.:Secrecy, authentication,

Accountable internet protocol (AIP. In: Proceedings of the ACM SIGCOMM 2008, pp. 339 350 (2007) 12.

ACM Transactions on Computer systems 2 (4), 277 288 (1984) 14. Lagutin, D.,Visala, K.,Tarkoma, S.:

Publish/Subscribe for Internet: PSIRP Perspective. Valencia FIA book (2010) 15. Tarkoma, S.,Antikainen, M.:

13th IEEE Global Internet Symposium 2010 (2010) 16. Gao, L.:On Inferring Autonomous System Relationships in the Internet.

IEEE/ACM Transactions on Networking 9 (6), 733 745 (2001) 17. Yang, X.,Clark, D.,Berger, A w.:

IEEE Computer Society Press, Los Alamitos (2004) 20. Carpenter, B.:rfc1958: Architectural Principles of the Internet.

IETF (June 1996) 21. Jokela, P.,Zahemszky, A.,Esteve, C.,Arianfar, S.,Nikander, P.:LIPSIN:

European Conference on Computer network Defence, EC2ND (2009) 23. Miller, V. S.:Use of elliptic curves in cryptography.

In: Williams, H. C. ed.)CRYPTO 1985. LNCS, vol. 218, pp. 417 426. Springer, Heidelberg (1986) 24.

Hardware subtask final report. Helsinki University of Technology, Tech. Rep (2008), http://www. tcs. hut. fi/Software/PLA/new/doc/PLA HW FINAL REPORT. pdf 25.

Lagutin, D.:Securing the Internet with Digital Signatures. Doctoral dissertation, Department of computer science and Engineering, Aalto University, School of Science and Technology (2010) Engineering Secure Future Internet Services Wouter Joosen1, Javier Lopez2, Fabio Martinelli3,

and Fabio Massacci4 1 Katholieke Universiteit Leuven wouter. joosen@cs. kuleuven. be 2 University of Malaga jlm@lcc. uma. es 3 National Research

Council of Italy Fabio. Martinelli@iit. cnr. it 4 University of Trento massacci@dit. unitn. it Abstract.

and the opportunity for establishing a discipline for engineering secure Future Internet Services, typically based on research in the areas of software engineering,

Generic solutions that ignore the characteristics of Future Internet services will fail, yet it seems obvious to build on best practices

It will be essential to integrate various activities that need to be addressed in the scope of secure service engineering into comprehensive software and service life cycle support.

in order to jointly enable the security and trustworthiness of Future Internet services. 1 Introduction 1. 1 Future Internet Services The concept named Future Internet (FI) aggregates many facets

The Future Internet may evolve to use new infrastructures, network technologies and protocols in support of a growing scale and a converging world, especially in light of smaller, portable, ubiquitous and pervasive devices.

Besides such a network-level evolution, the Future Internet will manifest itself to the broad mass of end users through a new generation of services (e g. a hybrid aggregation of content and functionality

) Future Internet Assembly, LNCS 6656, pp. 177 191,2011. c The Author (s). This article is published with open access at Springerlink. com. 178 W. Joosen et al. be operated

yet the Future Internet stretches the present know how on building secure software services and systems:

Furthermore, the Future Internet will be an intrinsically dynamic and evolving paradigm where, for instance, end users are empowered more and more

and reassessed continuously. 1. 2 The Need for Engineering Secure Software Services The need to organize,

integrate and optimize the research on engineering secure software services to deal effectively with this increased challenge is pertinent and well recognized by the research community and by the industrial one.

This obviously harms the economic impact of Future Internet services and causes significant monetary losses in recovering from those attacks.

however, we are now witnessing the emergence of new and unprecedented models for service-oriented computing for the Future Internet:

Infrastructure as a service (Iaas), Platform as a service (Paas) and Software as a service (Saas). These models have the potential to better adhere to an economy of scale

New Internet services will have to be Engineering Secure Future Internet Services 179 provided in the near future,

and damaged reputation. 1. 3 Research Focus on Developing Secure FI Services Our focus is on the creation and correct execution of a set of methodologies, processes and tools for secure software development.

approving that the developed software is secure. Assurance must be based on justifiable evidence, and the whole process designed for assurance.

This would allow the uptake of new ICT-services according to the latest Future Internet paradigms,

integrating the former results in (5) a risk-aware and cost-aware software development life-cycle (SDLC),

and (6) the delivery of case studies of future internet application scenarios. The first three activities represent major and traditional stages of (secure) software development:

from requirements over architecture and design to the composition and/or programming of working solutions.

These three activities interact to ensure the integration between the methods and techniques that are proposed

and techniques that we consider useful for engineering secure Future internet services. 2 Security Requirements Engineering The main focus of this research strand is to enable the modeling of high-level requirements that can be expressed in terms of

The need for assurance in the future Internet demands a set of novel engineering methodologies to guarantee secure system behavior and provide credible evidence that the identified security requirements have been met from the point of view of all stakeholders.

The security requirements of Future Internet applications will differ considerably from those of traditional applications.

The reason is that Future Internet applications will not only be distributed geographically as are traditional applications,

and may involve an array of physical devices such as smart cards, phones, RFID sensors and so on that are connected perpetually

and transmit a variety of information including identity, bank accounts, location, and so on. Some of these transactions might even happen transparently to the user;

Engineering Secure Future Internet Services 181 This picture is complicated further by the vast number and the geographical spread of smart devices stakeholders would deploy to meet their requirements.

Sensor networks, RFID tags, smart appliances that communicate not only with the user but with their manufacturers, are examples of such devices.

Such deployments inherit security risks from the classical Internet and, at the same time create new and more complex security challenges.

Examples include illicit tracking of RFID tags (privacy violation) and cloning of data on RFID tags (identity theft).

Applications that involve such deployments typically cross organization boundaries. In light of the challenges and principles highlighted above,

and methodologies for software construction as well as researching about new ways to take this complexity into account in a holistic manner.

The design phase of the software service and/or system is a timely moment to enforce

The software architecture encompasses the more relevant elements of the application, providing either a static or/and a dynamic view of the application.

which comprise software elements, the externally visible properties of those elements, and the relationships among them. 182 W. Joosen et al.

assess and reason about security mechanisms at an early phase in the software development cycle. The research topics one must focus on in this subarea relate to model-driven architecture and security, the compositionality of design models and the study of design patterns for FI services and applications.

Until this point in the software and service development process, different concerns security among them of the whole application have been separated into different models,

Engineering Secure Future Internet Services 183 in order to grasp a comprehensive understanding of the application as a whole,

A design pattern is a general repeatable solution to a commonly occurring problem in software design.

both from a general perspective and from a security perspective for security-critical software systems. 4 Security Support in Programming Environments Security Support in Programming Environments is not new;

still it remains a grand challenge, especially in the context of Future Internet (FI) Services.

Securing Future Internet Service is inherently a matter of secure software and systems. The context of the future internet services sets the scene in the sense that (1) specific service architectures will be used,

that (2) new types of environments will be exploited, ranging from small embedded devices (things) to service infrastructures and platform in the cloud,

and (3) a broad range of programming technologies will be used to develop the actual software and systems.

The search for security support in programming environments has to take this context in account.

The requirements and architectural blueprints that will be produced in earlier stages of the software engineering process cannot deliver the expected security value

Some of these properties have been embedded in the security specific elements of the software design; other may simply be high priority security requirements that have articulated such as the appropriate treatment of concurrency control and the avoidance of race conditions in the code,

Supporting security requirements in the programming code level requires a comprehensive approach. The service creation means must be improved

as well as programming new services from scratch using a state-of-the-art programming language. The service creation context will typically aim for techniques

and the objectives of community wide research activities. 4. 1 Secure Service Composition Future Internet services

Middleware Aspects. The research community should re-investigate service-oriented middleware for the Future Internet

with a special emphasis on Engineering Secure Future Internet Services 185 enabling deployment, access, discovery and composition of pervasive services offered by resource-constrained nodes. 4. 2

Secure Service Programming Many security vulnerabilities arise from programming errors that allow an exploit. Future Internet will further reinforce the prominence of highly distributed and concurrent applications,

making it important to develop methodologies that ensure that no security hole arises from implementations that exploit the computational infrastructure of the Future Internet.

The research community must further investigate advances over state-of-the-art in fine-grained concurrency to enable highly concurrent services of the Future Internet

and will improve analysis and verification techniques to verify, among others, adherence to programming principles and best-practices 10.

Verifiable Concurrency. Lock-free wait-free algorithms for common software abstractions (queues, bags, etc. are one of the most effective approaches to exploit multi-core parallelism.

These algorithms are hard to design and prove correct, error-prone to program, and challenging to debug.

Their correctness is crucial to the correct behaviour of client programs. Research should now focus on build independently checkable proofs of the absence of common errors,

Adherence to Programming Principles and Best-Practices. Programming support must include methods to ensure the adherence of a particular program to well-known programming principles or best-practices in secure software development.

Emphasis will be put on language extensions that guarantee adherence to best-practices and verified design patterns that can be used during development.

in order to prevent cross-site scripting attacks and similar vulnerabilities associated with web-based distributed applications.

Obviously, the logical rationales underlying such best-practises must be articulated, enabling he development of type systems enforcing these practises directly

while still maintaining security. 4. 3 Platform Support for Security Enforcement Future Internet applications span multiple trust domains,

Web technology inherently embeds the concept of cross-domain references and applications are isolated via the Same-Origin-Policy (SOP) in the browser.

Trustworthy applications need run-time execution monitors that can provably enforce advanced security policies 19,3 including fined-grained access control policies usage control policies

Assurance will play a central role in the development of software based services to provide confidence about the desired security level.

seamlessly informing and giving feedback at each stage of the software life cycle by checking that the related models

Obviously the security support in programming environments that must be delivered will be essential to incept a transverse methodology that enables to manage assurance throughout the software and service development life cycle (SDLC.

The next section clarifies these issues. 5 Embedding Security Assurance and Risk management during SDLC Engineering secure Future Internet services demands for at least two traversal issues,

security assurance and risk and cost management during SDLC. 5. 1 Security Assurance The main objective is to enable assurance in the development of software based services to ensure confidence about their trustworthiness.

Our core goal is to incept a transverse methodology that enables to manage assurance throughout the software development life cycle (SDLC.

Early detection of security failures in Future Internet applications reduces development costs and improves assurance in the final system.

by developing refinement strategies, from policies down to mechanisms, for more complex Engineering Secure Future Internet Services 187 secure protocols, services, and systems.

such as the AVISPA 1 tool set and the Scyther tool 7, for the verification of Future Internet protocols.

Security policies can be implemented correctly by construction through a rigorous secure programming discipline. Internet applications can be validated through testing.

In that case, it is possible to develop test data generation that specifically targets the integration of services

access control policies or specific attacks. Moreover, implementations can be monitored at run-time to ensure that they satisfy the required security properties.

Complementing activities are related to secure programming. This strand addresses a comprehensive solution for program verification,

We can consider three aspects, that although not comprehensive, present characteristic for service-oriented applications in the future Internet:

penetration testing that leverages on the high-level models that are generated in early stages of the software life cycle,

automated generation in XML-based input data to maximize the efficiency in the security testing process,

run-time verification must complement programming-level verification and testing in order to provide the final assurance that the latter cannot deliver,

typical for service compositions in Future Internet. We will study approaches for run-time monitoring of data flow,

as well as technologies for privacy-preserving usage control. Towards a Traverse Methodology. Security concerns are specified at the business-level

Metrics can be used directly for computing risks (e g.,, probability of threat occurrence) or indirectly (e g.,

Security metrics in the future Internet applications become increasingly important. Service-oriented architectures demand for assurance indicators that can explicitly indicate the quality of protection of a service,

Clients want to be sure that their data outsourced to other domains, which the clients cannot control,

and cost aware SDLC should be based on an incremental and iterative process that is accommodated to an incremental software development process.

While the software development proceeds through incremental phases, the risk and cost analysis will undergo new iterations for each phase.

and cost analyses will propagate through the software development phases and become more refined. In order to support the propagation of analysis results through the phases of the SDLC Engineering Secure Future Internet Services 189 one needs to develop methods and techniques for the refinement of risk analysis documentation.

Such refinement can be obtained both by refining the risk models e g. by detailing the description of relevant threats and vulnerabilities,

In order to accommodate to a modular software development process, as well as effectively handling the heterogeneous and compositional nature of Future Internet services,

one needs to focus on a modular approach to the analysis of risks and costs. In a compositional setting, also risks become compositional

secure programming as well as assurance and the relation to each of these ingredients must be investigated. During security requirements engineering risk analysis facilitates the identification of relevant requirements.

and the opportunity for firmly establishing a discipline for engineering secure Future Internet Services, typically based on research in the areas of software engineering, security engineering and of service engineering.

We have clarified why generic solutions that ignore the characteristics of Future Internet services will fail:

Work partially supported by EU FP7-ICT project NESSOS (Network of Excellence on Engineering Secure Future Internet Software Services and Systems) under the grant agreement n. 256980.

Software Architecture In practice, 2nd edn. Addison-Wesley, Boston (2003) 3. Bauer, L.,Ligatti, J.,Walker, D.:

An agent-oriented software development methodology. Autonomous Agents and Multi-Agent Systems 8, 203 236 (2004) 6. Clavel, M.,da Silva, V.,de O. Braga, C.,Egea, M.:

Springer, Heidelberg (2008) Engineering Secure Future Internet Services 191 8. Dardenne, A.,van Lamsweerde, A.,Fickas, S.:

Proceedings of the 22nd Annual Symposium on Foundations of Computer science, WASHINGTON DC, USA, pp. 350 357.

IEEE Computer Society Press, Los Alamitos (1981), doi: 10.1109/SFCS. 1981.32 10. Erlingsson, U.,Schneider, F. B.:

Irm enforcement of java stack inspection. In: Proceedings of the 2000 IEEE Symposium on Security and Privacy, WASHINGTON DC, USA, pp. 246 255.

IEEE Computer Society Press, Los Alamitos (2000) 11. France, R.,Fleurey, F.,Reddy, R.,Baudry, B.,Ghosh, S.:

Proceedings of the 11th IEEE International Enterprise Distributed Object Computing Conference, WASHINGTON DC, USA, p. 253.

IEEE Computer Society Press, Los Alamitos (2007) 12. Giorgini, P.,Mouratidis, H.,Zannone, N.:Modelling security and trust with secure tropos.

Integrating Security and Software engineering: Advances and Future Vision, IDEA (2006) 13. Group, O.:Security design pattern technical guide, http://www. opengroup. org/security/gsp. htm 14.

Proc. of the Workshop on Ubiquitous Knowledge discovery for Users at ECML/PKDD, pp. 51 64 (2006) 15.

A programming model for concurrent object-oriented programs. ACM Trans. Program. Lang. Syst. 31,1 1 (2008), doi:

SPAQU'08 (Int. Workshop on Software Patterns and Quality)( 2008) 18. Lazouski, A.,Martinelli, F.,Mori, P.:

Computer science Review 4 (2), 81 99 (2010) 19. Le Guernic, G.,Banerjee, A.,Jensen, T.,Schmidt, D. A.:

Security services architecture for secure mobile grid systems. Journal of Systems Architecture. In Press (2010) 24.

Progress in Informatics 5, 35 47 (2008) Towards Formal Validation of Trust and Security in the Internet of Services Roberto Carbone1, Marius Minea2, Sebastian Alexander M odersheim3

The formal verification of trust and security of the Internet of Services will significantly boost its development

and public acceptance. 1 Introduction The vision of the Internet of Services (Ios) entails a major paradigm shift in the way ICT systems

they are no longer the result of programming components in the traditional meaning but are built by composing services that are distributed over the network

In the Ios, services are business functionalities that are designed and implemented by producers, deployed by providers,

However, the new opportunities opened by the Ios will only materialize if concepts, techniques and tools are provided to ensure security.

) Future Internet Assembly, LNCS 6656, pp. 193 207,2011. c The Author (s). This article is published with open access at Springerlink. com. 194 R. Carbone et al.

and associated exploits that are already plaguing complex web-based security-sensitive applications, and thus severely affect the development of the future internet.

Moreover, security validation should be carried out at all phases of the service development process, in particular during the design phase by the service designers themselves or by security analysts that support them in their complex tasks,

thereby significantly improving the all-round security of the Ios. In this chapter, we give a brief overview of the main scientific and industrial challenges for such verification tools,

and public acceptance of the Ios. We proceed as follows. In Sections 2 and 3, we discuss, respectively,

Towards Formal Validation of Trust and Security in the Internet of Services 195 Second, SOAS are also distributed systems,

and authentication/integrity of the communicated data. More elaborate goals are structural properties (which can sometimes be reduced to confidentiality and authentication goals) such as authorization (with respect to a policy), separation or binding of duty,

for a given web service for online shopping one may require that every order will eventually be processed

we may require a separation of duty property, namely that for privacy Towards Formal Validation of Trust and Security in the Internet of Services 197 purposes,

the inherent complexity (heterogeneity, distribution and dynamicity) of the Internet of Services, the challenge of validating services and service-oriented applications cannot be addressed simply by scaling up the current generation of formal analysis approaches and tools.

Two key approaches for composing web services have been considered, which differ by their architecture: orchestration is centralized

and all web services can communicate directly. 198 R. Carbone et al. Several orchestration notions have been advocated (see, e g.,

However, in inter-organizational business processes it is crucial to protect sensitive data of each organization;

and our main motivation is to take into account the security policies while computing an orchestration. The AVANTSSAR Platform, for example, implements an idea presented in 11 to automatically generate a mediator.

We specify a web service profile from its XML Schema and WSSECURITYPOLICY using first-order terms (including cryptographic functions).

The mediator is able to use cryptography to produce new messages, and is constructed with respect to security goals using the techniques we developed for the verification of security protocols. 3. 2 Model Checking of SOAS Model checking 13 is a powerful and automatic technique for verifying concurrent systems.

It is, Towards Formal Validation of Trust and Security in the Internet of Services 199 of course,

For instance, Tulafale 6, a tool by Microsoft Research based on Proverif 7, exploits abstract interpretation for verification of web services that use SOAP messaging, using logical predicates to relate the concrete

or cryptographic keys) into finitely many equivalence classes and to compute on those equivalence classes instead;(

This allows for the use of classical automated first-order reasoning techniques, in particular resolution or fixed-point computations of static analysis. Thanks to the over-approximation,

The idea is to organize data by means of sets and to abstract data by set membership.

, a layer of software modules that carry out the translation from application-level specification languages (such as BPMN and BPEL,

Towards Formal Validation of Trust and Security in the Internet of Services 201 Vulnerability: Policy:

These must contain all relevant information required to determine the access to private data and to the meta-policies that control them.

making clinical and nonclinical data available anywhere and anytime in a health care organization, while lowering infrastructure costs.

A highlight of the effectiveness of the AVANTSSAR methods and tools is the detection of a serious flaw in the SAML-based SSO solution for Google Apps 3. Though well specified and thoroughly documented,

it is hard to establish which message fields are mandatory in a given Towards Formal Validation of Trust and Security in the Internet of Services 203 profile and

Still, when Google developed their SAML-based SSO solution for Google Apps they released a flawed product,

which allowed a dishonest service provider to impersonate the victim user on Google Apps, granting unauthorized access to private data and services (email, docs, etc.).

The vulnerability was detected by the SATMC backend of the AVANTSSAR Platform and the attack was reproduced in an actual deployment of SAML-based SSO for Google Apps.

Google and the US Computer Emergency Readiness Team (US-CERT) were informed and the vulnerability was kept confidential until Google developed a new version of the authentication service

and Google's customers updated their applications accordingly. The severity of the vulnerability has been rated High in a note issued by the National Institute of Standard and Technology (NIST.

Moreover, as shown in 2, the SATMC backend of the AVANTSSAR Platform also allowed us to detect that the prototypical SAML SSO use case (as described in the SAML technical overview) suffers from an authentication flaw that,

under some conditions, allows a malicious service provider to hijack a client authentication attempt and force the latter to access a resource without its consent or intention.

It also allows an attacker to launch Cross-Site Scripting (XSS) and Cross-Site Request Forgery attacks (XSRF.

as witnessed by the new XSS attack identified in the SAML-based SSO for Google Apps

and that could have allowed a malicious web server to impersonate a user on any Google application.

PKCS#11 specifies an API for performing cryptographic operations such as encryption and signature using cryptographic tokens (e g.,

, USB tokens or smart cards. Sensitive cryptographic keys, stored inside the token, should not be revealed to the outside

and it should be impossible for an attacker to change those keys. The attacks found show that in many implementations this is not the case:

Formal validation of trust and security will become a reality in the Internet of Services

Though the use of FM would promote a more secure development environment, a variety of practical and cultural reasons lead the industrial world to perceive FM approaches as being expensive in terms of time

and (iii) the differences between formal languages and models and those used in industrial design and development environments (e g.,

, BPMN, Java, ABAP. The problem is how to make new, efficient methodologies and technologies accessible and readily exploitable, benefitting industry designers and developers.

and translators to and from the core formal models should be devised and migrated to the selected development environments.

A concrete example is the industry migration of the AVANTSSAR Platform to the SAP environment.

Two valuable migration activities have been carried out by building contacts with core business units. First, in the trail of the successful analysis of Google's SAML-based SSO, an internal project has been run to migrate AVANTSSAR results within SAP Netweaver Security

and Identity Management (SAP NW SIM) with the objective of exploiting the AVANTSSAR technology to initiate a deep formal analysis of the SAP Netweaver SAML Next Generation Single Sign-on (NW-NGSSO

) to formally establish its soundness, i e.,, to have formal evidence that the employed service providers

and Security in the Internet of Services 205 there and helped SAP Research to better understand the vulnerability itself

and Outlook As exemplified by these case studies and success stories, formal validation technologies can have a decisive impact for the trust

and security of the Ios. The research innovation put forth by AVANTSSAR aims at ensuring global security of dynamically composed services

These advances will significantly improve the all-round security of the Ios, and thus boost its development and public acceptance.

Formal Analysis of SAML 2. 0 Web browser Single Sign-on: Breaking the SAML-based Single Sign-on for Google Apps.

In: Proceedings of the 6th ACM Workshop on Formal Methods in Security Engineering (FMSE 2008), pp. 1 10.

A security tool for web services. In: de Boer, F. S.,Bonsangue, M m.,, Graf, S.,de Roever, W.-P. eds.

IEEE Computer Society Press, Los Alamitos (2001) 8. Bodei, C.,Buchholtz, M.,Degano, P.,Nielson, F.,Nielson, H r.:

Proceedings of the 17th ACM conference on Computer and Communications security (CCS 2010), pp. 260 269.

IEEE Computer Society Press, Los Alamitos (2008) 12. Ciob aca, S.,Cortier, V.:Protocol composition for arbitrary primitives.

IEEE Computer Society Press, Los Alamitos (2010) 13. Clarke, E. M.,Grumberg, O.,Peled, D. A.:

Proceedings of the 19th MFPS, ENTCS 83, Elsevier, Amsterdam (2004) Towards Formal Validation of Trust and Security in the Internet of Services 207 17.

Journal of Logic and Algebraic Programming 70 (1), 96 118 (2007) 20. Marconi, A.,Pistore, M.:

Abstraction by Set-Membership Verifying Security Protocols and Web Services with Databases. In: Proceedings of 17th ACM conference on Computer and Communications security (CCS 2010), pp. 351 360.

ACM Press, New york (2010) 22. M odersheim, S.,Vigan`o, L.:Secure Pseudonymous Channels. In:

Web Services Business Process Execution Language vers. 2. 0 (2007), http://docs. oasis-open. org/wsbpel/2. 0/OS/wsbpel

-v2. 0-OS. pdf 25. Pnueli, A.:The Temporal Logic of Programs. In: Proceedings of the 18th IEEE Symposium on Foundations of Computer science, pp. 46 57.

IEEE Computer Society Press, Los Alamitos (1977) 26. T. Dierks and E. Rescorla. The Transport Layer Security (TLS) Protocol, Version 1. 2. IETF RFC 5246 (Aug. 2008) 27.

Turuani, M.:The CL-Atse Protocol Analyser. In: Pfenning, F. ed.)RTA 2006. LNCS, vol. 4098, pp. 277 286.

Springer, Heidelberg (1999) Trustworthy Clouds Underpinning the Future Internet R udiger Glott1, Elmar Husmann2, Ahmad-Reza Sadeghi3,

and Matthias Schunter2 1 Maastricht University, The netherlands glott. ruediger@gmail. com 2 IBM Research Z urich, R uschlikon, Switzerland huselmar@de

. ibm. com, mts@zurich. ibm. com 3 TU Darmstadt, Germany ahmad. sadeghi@trust. rub. de Abstract.

Cloud computing is a new service delivery paradigm that aims to provide standardized services with self-service,

run-time platform, or actual Software as a service. They are are expected to be an important component in the future Internet.

This article introduces upcoming security challenges for cloud services such as multi-tenancy, transparency and establishing trust into correct operation,

and survey related research in these areas. 1 Cloud computing and the Future Internet Cloud computing is expected to become a backbone technology of the Future Internet that provides Internet-scale

and service-oriented access to virtualized computing, data storage and network resources as well as higher level services.

cloud computing in the future Internet is expected to be characterized by a seamless cloud capacity federation of independent providers-similar to the network peering

and IP transit purchasing of ISPS in today's Internet. For an end-user this means that via interacting with one cloud provider,

Cloud computing goes beyond technological infrastructure that derives from the convergence of computer server power, storage and network bandwidth.

and distribution model for computing that establishes a new relationship between the end user and the data center,

gives the user'programmatic control'over a part of the data center 1, pp. 8-9. For this cloud-of-clouds vision4this article will investigate the related challenges for trust

and security architectures and mechanisms. 4 For which the Internet pioneer Vint Cerf has suggested recently the term Intercloud J. Domingue et al.

) Future Internet Assembly, LNCS 6656, pp. 209 221,2011. c The Author (s). This article is published with open access at Springerlink. com. 210 R. Glott et al.

FIA projects like RESERVOIR or VISION are conducting research on core technological foundations of the cloud-of-clouds such as federation technologies, interoperability standards or placement policies for virtual images or data

Many of these developments can be expected to be transferred into the Future Internet Core Platform project that will launch in 2011.

the SNIA Cloud storage Technical Working group or the OGF Open Clouds Computing Interface Working group. Trust and security are regarded often as an afterthought in this context,

Today, since the current legal systems are prepared not for the challenges that result from the complexity and pervasiveness of cloud computing,

data protection and privacy issues as well as liability and compliance problems may hinder to tap the full potential of cloud computing 22,

in the sense that it will ensure that data mobility is limited to ensure compliance with a wide range of different national legislation including privacy legislation such as the EU Data protection Directive 95/46/EC.

As of today, cloud computing is facing significant acceptance hurdles when it comes to hosting important business applications

Trustworthy Clouds Underpinning the Future Internet 211 An example for the first category is the Google gov. app cloud launched in September 2009 that offers a completely segregated cloud targeted exclusively at US government customers.

Similarly, IBM has launched a FISMA compliant Federal Community Cloud in 2010. Other cloud providers also adapt basic service security to the needs of specific markets and communities.

Following its software-plus-services strategy announced in 2007 Microsoft has developed in the past years several Saas cloud services such as the Business Productivity Online Suite (BPOS.

While all of them may be delivered from a multi-tenant public cloud for the entry level user, Microsoft offers dedicated private cloud hosting

and supports third-party or customer-site hosting. This allows tailor made solutions to specific security concerns-in particular in view of the needs of larger customers.

In the same way, the base security of Microsoft public cloud services is adapted to the targeted market.

Whereas Microsoft uses, e g.,, for the Office Live Workspace-in analogy to what Google does with Gmail-unencrypted data transfer between the cloud and the user

cloud services for more sensitive markets (such as Microsoft Health Vault) use SSL encryption by Default on the other hand commodity public cloud services such as the Amazon EC2 are still growing

even though they offer only limited base security and largely transfer responsibility for security to the customer.

, Novell, IBM), virtual private networking (e g.,, Amazon Virtual Private cloud), encryption (e g.,, Amazon managed encryption services)

and web traffic filtering services (e g.,, Zscaler, Scansafe. 2. 2 Today's Datacenters as the Benchmark for the Cloud Using technology always constitutes a certain risk.

, firewalls, intrusion defense, and protection of each host), all systems usually contain errors that can be exploited found

exchanging media such as USB STICKS allows transfer into systems that are connected not to the Internet 5. Cloudsourcing 15 follows more or less the same economic rationale as traditional IT-outsourcing

inter alia with regard to upgrades and patches, quick procurement services, avoidance of vendor lock ins, and legacy modernization 18.

cloud computing might be hindered significantly by the legal problems that remain to be solved. For the security objectives when adopting clouds for hosting critical systems we believe that today's datacenters are the benchmark for new cloud deployments.

and Privacy Risks and Emerging Security Controls Cloud computing being a novel technology introduces new security risks 7 that need to be mitigated.

Security Risks 12) Trustworthy Clouds Underpinning the Future Internet 213 3. 1 Isolation Breach between Multiple Customers Cloud environments aim at efficiencies of scale by increased

As a consequence, data leakage and service disruptions gain importance and may propagate through such shared resources.

An important requirement is that data cannot leak between customers and that malfunction or misbehavior by one customer must not lead to violations of the service-level agreement of other customers.

and data wiping before reuse. Sharing of resources and multi-tenant isolation can be implemented on different levels of abstraction (see Figure 2). Coarse-grained mechanisms such as shared datacenters, hosts,

and networks are understood well and technologies such as virtual machines, vlans, or SANS provide isolation. Sharing resources such as operating systems, middleware,

or actual software requires a case-by-case design of isolation mechanisms. In particular the last example of Software-as-a-service requires that each data instance is assigned to a customer

and that these instances cannot be accessed by other customers. Note that in practice, these mechanisms are mixed often:

While an enterprise customer may own a virtual machine (Machine-level isolation), this machine may use a database server (Middleware isolation)

and provide services to multiple individual departments (Application isolation). In order to mitigate this risk in a cloud computing environment,

multi-tenant isolation ensures customer isolation. A principle to structure isolation management is One way to implement such isolation is labeling

and flow control: Labeling: By default all resources are assigned to a customer and labeled with a corresponding label.

Shared resources must moderate potential data flow and ensure that no unauthorized data flow occurs between customers.

To limit flow control, mechanisms such as access control that ensures that machines and applications of one customer cannot access data

or resources from other customers can be used. Actual systems then need to implement this principle for all shared resources 4 (see, e g.,

Examples may include a network administrator impacting database operations or administrators stealing and disclosing data.

Customer employees can access their respective data and systems (or parts thereof) but cannot access infrastructure

or data owned by different customers. This so-called privileged identity management system is starting to be implemented today

For instance, a database administrator may only obtain administrative privileges over the tables owned by its employer. 2. For a given task at hand,

a database administrator asks for privileges to modify a given database schema. 3. The administrator performs the desired task. 4. The administrator returns the privileges.

, trusted computing 21 or computations on outsourced data 20. Trustworthy Clouds Underpinning the Future Internet 215 3. 3 Failures of the Cloud Management Systems Due to the highly automated nature of the cloud management systems

and the high complexity of the managed systems, software quality plays an important role in avoiding disruptions and service outages:

Clouds gain efficiency by industrializing the production of IT services through complete end-to-end automation. This means that once errors occur in such complex and automated systems,

Another source of failure stems from the fact that large-scale computing clouds are built often using low-cost commodity hardware that fails (relatively) often.

The consequence of these facts is automated that fault tolerance problemdetermination, and (self-)repair mechanisms will be needed commonly in the cloud environment

or recover from software and hardware failures. For building such resilient systems, important tools are data replication,

atomic updates of replicated management data, and integrity checking of all data received (see, e g.,

, 24. In the longer run, usage of multiple clouds may further improve resiliency (e g.,, as pursued by the TCLOUDS project www. tclouds-pro ject. eu or proposed in 11). 3. 4 Lack of Transparency

and Guarantees While the proposed mechanisms to mitigate the identified risks are important, security incidents are largely invisible to a customer:

Data corruption may not be detected for a long time. Data leakage by skilled insiders is unlikely to be detected. Furthermore, the operational state and potential problems are communicated usually not to the customer except after an outage has occurred.

An important requirement in a cloud setting is to move away from today's black-box approach to cloud computing where customers cannot obtain insight on or evidence of correct cloud operations.

A related challenge is how to best foster trust of customers into correct operation of the cloud infrastructure.

and no data is corrupted or leaked. In practice, these problems are unsolved largely. Cryptographers have designed schemes such as homomorphic encryption 9 that allow verifiable computation on encrypted data.

However, the proposed schemes are too inefficient and do not meet the complete range of privacy requirements 23.

A more practical solution is to use Trusted Computing to verify correct policy enforcement 6. Trusted computing instantiation as proposed by the Trusted Computing Group (TCG) uses secure hardware to allow a stakeholder

To enable trusted cloud computing, privacy protection is an essential requirement 26. In simple terms, data privacy aims at protecting personally identifiable data (PID.

In Europe, Article 8 of the European Convention on Human rights (ECHR) provides a right to respect for ones private and family life, his home and his correspondence.

Furthermore, the European Data protection Directive (Directive 95/46/EC) substantiates this right in order to establish a comprehensive data protection system throughout Europe.

This directive takes into account the OECD privacy principles 16 which mandate several principles such as, e g.,

, limited collection of data the authorization to collect data either by law or by informed consent of the individual whose data are processed (data subject),

the right to correction and deletion as well as the necessity of reasonable security safeguards for the collected data.

Since cloud computing often means outsourcing data processing, the user as well as the data subject might face risks of data loss,

corruption or wiretapping due to the transfer to an external cloud provider. Related to these de-facto obstructions in regard to the legal requirements, there are three particular challenges that need to be addressed by all cloud solutions:

Transparency, technical and organizational security safeguards and contractual commitments (e g.,, Service Level Agreements, Binding Corporate Rules.

According to European law, the user who processes PID in the cloud or elsewhere remains responsible for the compliance with the aforementioned principles of data privacy.

Outsourcing data processing does not absolve the user from his responsibilities and liabilities concerning the data.

This means that the user must be able to control and comprehend what happens to the data in the cloud and

which security measures are deployed. Therefore, the utmost transparency Trustworthy Clouds Underpinning the Future Internet 217 regarding the processes within the cloud is required to enable the user to carry out his legal obligations.

This might be realized technically by e g.,, installing informative event and access logs which enable the user to retrace in detail what happens to his data,

where they are stored and who accesses them. Also, the cloud service provider could prove to have an appropriate level of security measurements by undergoing acknowledged auditing

This applies all the more in cases of cross-border cloud computing with various subcontracting cloud service providers.

Subcontracts are practiced already commonly in the cloud computing field. Cloud services commonly rely on each other, since their structures may be based consecutively upon each other.

Hence, a computing cloud may use the services of a storage cloud. Unlike local data centers residing in a single country

such cloud infrastructures often extend over multiple legislation and countries. Therefore, the question of applicable law and safeguarding the user's responsibilities regarding data privacy in cross-border cloud scenarios is a matter of consequences for the use of these cloud services.

So to avoid unwanted disclosure of data, sufficient protection mechanisms need to be established. These may also extend to the level of technical solutions, such as encryption,

data minimization or enforcement of processing according to predefined policies. 4 Open Research Challenges Today's technology for outsourcing

and large-scale systems management laid the foundation for cloud computing. Nevertheless, due to its global scale and the need for full automation, there are still open research challenges that need to be resolved

in order to enable hosting of enterprise-class and critical systems on a cloud. Customer Isolation and Information Flow.

Furthermore, data generated by systems need to be assigned to one or more customers to enable access to critical data such as logs and monitoring data.

A particularly hard challenge will be to reduce the amount of covert and side channels. Today

such channels are frozen often in hardware and thus cannot easily be reduced. 218 R. Glott et al.

Today, regulations often mandate that data needs to be processed in a particular country. This does not align well with today's cloud architectures

Trustworthy Clouds Underpinning the Future Internet 219 5 Outlook The Path Ahead Cloud computing is not new it constitutes a new outsourcing delivery model that aims to be closer to the vision of true utility computing.

and data integrity through authentication. However, we expect that they will then move on to the harder problems such as providing verifiable transparency,

Proceedings of the 2010 IEEE 30th International Conference on Distributed computing Systems Workshops, WASHINGTON DC, USA. ICDCSW'10, pp. 77 81.

IEEE Computer Society Press, Los Alamitos (2010), doi: 10.1109/ICDCSW. 2010.39 4. Cabuk, S.,Dalton, C i.,Eriksson, K.,Kuhlmann, D.,Ramasamy, H. V.,Ramunno, G.,Sadeghi, A r.,Schunter

Towards automated security policy enforcement in multi-tenant virtual data centers. J. Comput. Secur. 18,89 121 (2010) 220 R. Glott et al. 5. Chien, E.:

From http://www. symantec. com/connect/blogs/w32stuxnet-dossier 6. Chow, R.,Golle, P.,Jakobsson, M.,Shi, E.,Staddon, J.,Masuoka, R

Controlling data in the cloud: outsourcing computation without outsourcing control. In: ACM Workshop on Cloud computing Security (CCSW'09), pp. 85 90.

ACM Press, New york (2009) 7. Cloud Security Alliance (CSA: Top threats to cloud computing, version 1. 0. March 2010), http://www. cloudsecurityalliance. org/topthreats/csathreats. v1. 0. pdf 8. Computer and Communication

Industry Association (CCIA: Cloud computing (2009), http://www. ccianet. org/CCIA/files/cclibraryfiles/Filename/000000000151/Cloud computing. pdf 9. Gentry, C.:

Fully homomorphic encryption using ideal lattices. In: Proceedings of the 41st annual ACM symposium on Theory of computing, Bethesda, MD, USA.

STOC'09, pp. 169 178. ACM Press, New york (2009), doi: 10.1145/1536414.1536440 10. Grobauer, B.,Schreck, T.:

Towards incident handling in the cloud: challenges and approaches. In: Proceedings of the 2010 ACM workshop on Cloud computing security workshop, Chicago, Illinois, USA.

CCSW'10, pp. 77 86. ACM Press, New york (2010), doi: 10.1145/1866835.1866850 11. Guerraoui, R.,Yabandeh, M.:

Independent faults in the cloud. In: Proceedings of the 4th International Workshop on Large scale Distributed systems and Middleware, Z urich, Switzerland.

LADIS'10, pp. 12 17. ACM Press, New york (2010), doi: 10.1145/1859184.1859188 12. International organization for Standardization (ISO:

Proceedings of the 2nd USENIX conference on Hot topics in cloud computing. pp. 13 13. Hotcloud'10, USENIX Association, Berkeley, CA, USA (2010), http://portal. acm. org/citation. cfm?

& service brokers (2010), http://www. processor. com/editorial/article. asp? article=articles%2fp3203%2f39p03%2f39p03. asp 15.

Guidelines on the protection of privacy and transborder flows of personal data. From http://www. oecd. org/document/18/0, 2340, en 2649 34255 1815186 1 1 1 1, 00. html (last modified January 5 1999), the OECD Privacy Principles 17.

Proceedings of the 16th ACM conference on Computer and communications security, Chicago, Illinois, USA. CCS'09, pp. 199 212.

10.1145/1653662.1653687 Trustworthy Clouds Underpinning the Future Internet 221 20. Sadeghi, A r.,Schneider, T.,Winandy, M.:

Token-Based Cloud computing Secure Outsourcing of Data and Arbitrary Computations with Lower Latency. In: Acquisti, A.,Smith, S.,Sadeghi, A r. eds.

Proceedings of the 3rd international conference on Trust and trustworthy computing, Berlin, Germany, June 21-23,2010.

Towards trusted cloud computing. In: Proceedings of the 2009 conference on Hot topics in cloud computing. pp. 3 3. Hotcloud'09, USENIX Association, Berkeley, CA, USA (2009), http://portal. acm

. org/citation. cfm? id=1855533.1855536 22. Sotto, L j.,Treacy, B c.,Mclellan, M. L.:Privacy and data security risks in cloud computing. Electronic commerce & Law Report 15,186 (2010) 23.

Van dijk, M.,Juels, A.:On the Impossibility of Cryptography Alone for Privacy-Preserving Cloud computing. IACR eprint 305 (2010) 24.

Vukoli'c, M.:The byzantine empire in the intercloud. SIGACT News 41,105 111 (2010), doi: 10.1145/1855118.1855137 25.

Cloud computing and security. Lecture Univ. Stuttgart (November 2009) 26. Weichert, T.:Cloud computing und Datenschutz (2009), http://www. datenschutzzentrum. de/cloud-computing/Data Usage Control in the future Internet Cloud Michele Bezzi and Slim Trabelsi SAP Labs

, 06253, Mougins, France Abstract. The increasing collection of private information from individuals is becoming a very sensitive issue for citizens, organizations, and regulators.

in order to try to control the terms of usage of these collected data, but generally not providing a real efficient solution.

the data owners and the data collectors to verify the compliance of the data usage conditions with the regulations.

Recent studies address these issues by proposing a policy-based framework to express data handling conditions

and enforce the restrictions and obligations related to the data usage. In this paper, we first review recent research findings in this area, outlining the current challenges.

and visualize the use of their data stored in a remote server or in the cloud.

which monitors and informs the user on the compliance with a previously agreed privacy policy.

Privacy, Usage control, Privacy Policy 1 Introduction The vision of the Future Internet heralds a new environment where users,

In the cloud users and businesses can buy computing resources (e g.,, servers, services, applications) provided by the cloud

that are provisioned rapidly with a minimal management effort and pay-peruse. In the cloud, data may flow around the world,

ignoring borders, across multiple services, all in total transparency for the user. However, this ideal cloud world raises concerns about privacy for individuals, organizations,

In fact, when data cross borders, they have to comply with privacy laws in every jurisdiction,

and every jurisdiction has its own data protection laws. In addition, the risk, for personal data to travel across boundaries

and business domains, is that the usage conditions agreed J. Domingue et al. Eds.):) Future Internet Assembly, LNCS 6656, pp. 223 231,2011. c The Author (s). This article is published with open access at Springerlink. com. 224 M. Bezzi

and S. Trabelsi upon collection are lost, and, as a consequence, users cannot control their personal information any more,

as well as, honest businesses may lose confidence in handling data, when usage conditions are uncertain. To face these challenges,

expressing that the data should be used for specific purposes only, or the retention period should not exceed 6 months,

when data are transfered to a third party). The sticky policy is propagated with the information throughout its lifetime,

and data processors along the supply chain of the cloud have to handle the data in accordance with their attached policies.

such as setting and comparing user preferences with server privacy policies, expressing conditions on complex secondary usage cases,

Providing the data owner with a user-friendly way to express their preferences, as well as to verify the privacy policy the data are collected with.

Develop mechanisms to enforce these sticky policies in ways that can be verified and audited. In this paper

and data handling policies; we then describe the corresponding policy engine, enabling the deployment, interpretation and enforcement of PPL policies.

the current framework lacks mechanisms to provide the data owner with the guarantee that policy

Conclusions are drawn in the last section. 2 Primelife Privacy Framework In many web applications users are asked to provide various kinds of personal information, starting from basic contact information (addresses, telephone, email) to more complex data such as preferences, friends'list, photos.

Service providers Data Usage Control in the future Internet Cloud 225 Fig. 1. PPL high level architecture. describe how the users'data are handled using privacy policy,

which is presented, more or less explicitly to users during the data collection phase. Privacy policies are composed typically of a long text written in legal terms that are understood rarely fully,

or even read, by the users. As a result most of the users creating accounts on web 2. 0 applications are not aware of the conditions under

which their data are handled. Therefore, there is need to support the user in this process, providing an as-automatic-as-possible means to handle privacy policies.

In this context, the European FP7 project Primelife1 developed a novel privacy policy framework able to express and automatically process privacy policies in web interactions.

This approach enables applications, like web browsers, to automate the interpretation of the content of a privacy policy

and to compare the service privacy policy with user privacy preferences. The Primelife project introduced the Primelife Policy Language (PPL

which allows to describe in an XML machine-readable format the conditions of access and usage of the data.

A PPL policy can be used by a service provider to describe his privacy policies (how the data collected will be treated and with

or by a user to specify his preferences about the use of his data (who can use it

the user can automatically match his preferences with the privacy policy of the website and the result of the matching generates an agreed policy,

which is bound to the data (sticky policy) and travels with them. In fact, this sticky policy will be sent to the server

and follow the data in all their lifecycle to specify the usage conditions. The PPL sticky policy defines the following conditions:

1 www. primelife. eu 226 M. Bezzi and S. Trabelsi Access control: PPL inherits from the XACML 8 language the access control capabilities that express how access to which resource under

Data Handling: the data handling part of the language defines two conditions: Purpose: expressing the purpose of usage of the data.

Purpose can be for example marketing, research, payment, delivery, etc. Downstream usage: supporting a multilevel nested policy describing the data handling conditions that are applicable for any third party collecting the data from the server.

This nested policy is applicable when a server storing personal data decides to share the data with a third party Obligations:

Obligations in sticky policies specify the actions that should be carried out after collecting or storing a data.

For example, notification to the user whenever his data are shared with a third party, or deleting the credit card number after the payment transaction is finished, etc..

Introducing PPL policies requires the design of a new framework for the processing of such privacy rules.

In particular, it is important to stress that during the lifecyle of personal data, the same actor may play the role of both data collector and data provider.

For this reason, Primelife proposed the PPL engine based on a symmetric architecture, where any data collector can become a data provider

if a third party requests some data (see Figure 1). According to the role played by an entity (data provider

or data collector) the engine behaves differently by invoking the appropriate modules. In more detail

on the data provider side (user) the modules invoked are: The access control engine: it checks if there is any access restriction for the data before sending it to any server.

For example, we can define black or white lists for websites with whom we do not want to exchange our personal information.

Policy matching engine: after verifying that a data collector is in the white list, a data provider recovers the server's privacy policy

in order to compare it to its preferences and verify whether they are compatible in terms of data handling and obligation conditions.

The result of this matching may be displayed through a graphical interface, where a user can clearly understand how the information is handled

if he accepts to continue the transaction with the data collector. The result of the matching conditions,

as agreed by the user, is transformed into a sticky policy. On the data collector side

after recovering the personal information with its sticky policy the invoked modules are: Event handler:

it monitors all the events related to the usage of the collected data. These event notifications are handled by the obligation engine

in order to check if there is any trigger that is related to an event. For example, if a sticky policy provides for the logging of any information related to the usage of a data,

the event handler will notify the obligation engine whenever an Data Usage Control in the future Internet Cloud 227 access (read,

write, modification, deletion etc.)to data is detected in order to keep track of this access. Obligation engine:

it triggers all the obligations required by the sticky policy. If a third party requests some data from the server,

the latter becomes a data provider and acts as a user-side engine invoking access control and matching modules,

and the third party plays the role of data collector invoking the obligation engine and the event handler 3 Open Challenges Although the PPL framework represents an important advancement in fulfilling many privacy requirements of the cloud scenario,

there are still some issues, which are addressed not by the PPL framework. Firstly, in the current PPL framework

the data owner has no guarantee of actual enforcement of the data handling policies and obligations.

Indeed, the data collector may implement the PPL framework, thus having the technical capacity of processing the data according to the attached policies,

but it could always tamper with this system, which controls, or simply access directly the data without using the PPL engine.

In practice, the data owner should trust the data collector to behave honestly. A second problem relates to the scalability of the sticky policy approach.

Clearly, the policy processing adds a relevant computational overhead. Its applicability to realistic scenarios where large amounts of data have to be transmitted

and processed, has to be investigated. A last issue relates to the privacy business model. The main question is:

What should motivate the data collectors/processors to implement such technology? Actually, in many cases, their business model relies on the as-less-restricted-aspossible use of private data.

On the user side, a related question is, are the data owners ready to pay for privacy 9?

Both questions are difficult to address, especially when dealing with such a loosely defined concept as privacy.

Although studies exist (see 11,3, and references therein), mainly in the context of the web 2. 0, we should notice that the advent of cloud changes the business relevance of privacy.

In fact, in a typical web 2. 0 application the user is disclosing his own data,

balancing the value of his personal data with the services obtained. As a matter of fact, users have difficulties to monetize the value of their personal information,

and they tend to disclose their data quite easily. In the cloud world organizations store the data they have collected (under specific restrictions) with the cloud provider.

These data have a clear business value, and typically companies can evaluate the amount of money they are risking

if such data are lost or made public. For these reasons, it is likely that they are ready to pay for a stronger privacy protection.

All these issues need further research work to be addressed. In the next section, we present our initial thoughts on how we may extend the Primelife framework to address the first problem we mentioned above, i e.,

, how to provide a secure enforcement for privacy policy. 228 M. Bezzi and S. Trabelsi Fig. 2. The key elements of the extension of the PPL framework to guarantee the enforcement of privacy policy. 4 Towards Privacy Policy Enforcement in the Cloud In the current PPL framework,

there is no guarantee of enforcement of the data handling policies and obligations. In other words, we suppose that the server enforces correctly the sticky policies,

but, actually, nothing prevents him from creating a back door in his database in order to get unauthorized access to the collected information.

For this reason, we propose in the rest of the paper a secure architecture for the enforcement of the sticky policies

as well as giving the user control on the released data. The main idea is to introduce tamperproof 6 obligation engine and event handler, certified by a trusted third party,

which mediate the communication and the handling of private data in the cloud platform. The schedule of the events,

or more services/applications provided by external parties that deal with personal data (e g.,, a human resource management application, a remote storage service.

Say, these services handle personal data using a PPL framework (as described in Sect. 2) . In order to guarantee enforcement of the privacy policies and corresponding obligations by the service,

and event handler Data Usage Control in the future Internet Cloud 229 with a tamper-proof event handler and a tamper-proof obligation engine certified by a trusted third party (e g.,

If the data owner has the guarantee from a trusted authority (governmental office EU commission, etc.

he will tend to transfer his data to the certified host. In order to certify the compliance of an application,

if the stored data are handled correctly. The difficulty comes for the access to the database by the service provider.

One solution would be to use a specific tamper-proof database, but this can be technically complex,

and impact the business efficiency of the service provider. A possible solution is to specify an API to access the database that is compatible with the event handler.

This API should be defined as a standard interface to communicate with the event handler and access to the database.

The service has to exclusively use an interface compatible with the standardized API, and this should be subject to audit by an external trust authority

(which could be the same or not certifying the tamper proof components). Fig. 3. A sketch of data track administration console The particularity of this API is that all the methods to access the data can be detected by the event handler.

For example, if the service adds a new element (data and sticky policy) this action should be detected,

managed and logged by the event handler. If there is any method (like table dump) to access the database that cannot be recognized by the event handler,

the service will not be certified by the trusted authority. Using a tamper proof event handler and obligation engine also gives the possibility of providing a monitoring console.

The monitoring can be accessible by any data owner who, once authenticated, can list all the data (or set of data) with their related events and pending or enforced obligations.

The data owner can at any time control how his data are handled, under which conditions the information is accessed,

and compare them with the corresponding stored sticky policy. Fig. 3 shows a very simple example of how the remote administrative console could be structured,

this monitoring console could of course be more complex. The remote monitoring console adds more transparency

and more control to the data hosted within the cloud. It also allows the user to detect any improper usage of his data

and, in this case, notify the host or the trusted authority. 230 M. Bezzi and S. Trabelsi The advantages of the proposed solution are twofold.

First, from the data owner perspective, there is a guarantee that actual enforcement has taken place, and that he can monitor the status of his data and corresponding policies.

Second, from the auditors'point of view, it limits the perimeter of their analysis, since the confidence zone provided by the tamper proof elements

and the standardized API facilitate the distinction between authorized and non authorized actions. 5 Conclusions Cloud computing

and the SOA paradigm are fundamental building blocks for the Future Internet, enabling the seamless combination of services across platforms, geographies, businesses and transparently from the user point of view.

However, these new capabilities may entail privacy risks. From the user perspective, the risk is that of losing control of his personal information once they are released in the cloud.

In particular, when personal data are consumed by multiple services, possibly owned by different entities in different locations, the conditions of the data usage,

agreed upon collection, may be lost in the lifecycle of the personal data. From the data consumer point of view, businesses and organizations seek to ensure compliance with the plethora of data protection regulations

and minimize the risk of violating the agreed privacy policy. The concept of sticky policy may be used to address some of the privacy requirements of the cloud scenario.

In this paper we reviewed the recently introduced PPL framework, which provides a flexible language to express privacy policy as well as the necessary mechanisms to process

it notably requires a high level of trust in the data collector/processor. We presented some initial thoughts about how this problem can be mitigated through the usage of a tamper proof implementation of the architecture.

Enterprise privacy authorization language (EPAL 1. 1). IBM Research Report (2003) Data Usage Control in the future Internet Cloud 231 3. Bonneau, J

on the market for data protection in social networks. In: Moore, T.,Pym, D.,Ioannidis, C. eds.

IEEE International Workshop on Policies for Distributed systems and Networks, pp. 22 29 (2010) 5. Karjoth, G.,Schunter, M.,Waidner, M.:

Privacy-enabled management of customer data. In: Dingledine, R.,Syverson, P. F. eds. PET 2002.

Trust and tamper-proof software delivery. In: Proceedings of the 2006 international workshop on Software engineering for secure systems.

SESS'06, New york, NY, USA, pp. 51 58. ACM Press, New york (2006), doi: 10.1145/1137627.1137636 7. Reagle, J.,Cranor, L. F.:

extensible access control markup language (xacml) version 3. 0, extensible access control markup language (xacml) version 3. 0, oasis (August 2008) 9. Shostack, A.,Syverson, P.:

W3c Workshop on Privacy and data usage control p. 5 october 2010), http://www. w3. org/2010/policy-ws/11.

Future Internet Foundations: Experiments and Experimental Design Part IV: Future Internet Foundations: Experiments and Experimental Design 235 Introduction Research into new paradigms and the comprehensive test facilities upon which the ideas are experimented upon together build a key resource for driving European research into future networks and services.

This environment enables both incremental and disruptive approaches, supports multi-disciplinary research that goes beyond network layers, scholastic dogmas and public-private discussions.

It provides a core infrastructure, and also a playground for future discoveries and innovations, combining research with experimentation.

The heterogeneous and modular field of Future Internet Research and Experimentation with its national and international stakeholder groups requires community and cohesion building

and Resource Allocation Algorithms on the Federated Environment of Panlab reports on experiments needing to directly interact with the environment during runtime,

and solutions for a significant upgrade of the federated testbed environment that was used. The chapter by Zseby et al. entitled Multipath Routing Experiments in Federated Testbeds demonstrates the practical usefulness of federation and virtualisation in heterogeneous testbeds.

These multipath routing slice experiments were performed over multiple federated testbeds offered by the G-Lab

Finally the chapter Kousaridas et al. entitled Testing End-to-end Self management in a Wireless Future Internet Environment reports on the network management protocol test that exploited the availability of different administrative domains in federated testbeds

Future Internet Assembly, LNCS 6656, pp. 237 245,2011. The Author (s). This article is published with open access at Springerlink. com. A Use-Case on Testing Adaptive Admission Control and Resource Allocation Algorithms on the Federated Environment of Panlab Christos

Tranoris, Pierpaolo Giacomin, and Spyros Denazis Electrical and Computer engineering department, University of Patras, Rio, Patras 26500, Greece tranoris@ece. upatras. gr, yrz@anche. no, sdena@upatras. gr

Abstract. Panlab is a Future Internet initiative which integrates distributed facilities in a federated manner.

Panlab framework provides the infrastructure and architectural components that enable testing applications near production environments over a heterogeneous pool of resources.

This paper presents a use case where an adaptive resource allocation algorithm was tested utilizing Panlab's infrastructure.

As a result of this use case a new feature for Panlab was developed called Federation Computing Interface (FCI) API

Panlab, experimental testing, resource federation, Future Internet 1 Introduction Future Internet research results in new experimental infrastructures for supporting approaches that exploit extend

or redesign current Internet architecture and protocols. The Pan-European laboratory 1, Panlab, is a FIRE 2 initiative

and Architecture Elements to be used for experimentation in the future Internet. The Panlab infrastructure manages interconnections of different geographically distributed testbeds to provide services to customers for various kinds of testing scenarios which in Panlab terminology are called Virtual Customer Testbeds or simply VCTS.

A Web portal is available where customers and providers can access services, a visual Creation Environment

A Resource Adapter (a concept similar to device drivers) wraps a domain's resource API in order to create a homogeneous API defined by Panlab.

Details and specifications of Panlab's components can be found at 1. This paper describes an experiment made utilizing the Panlab's framework and available infrastructure.

i) to run the experiment by moving a designed algorithm from a simulating environment to near production besteffort environment

As a result to accomplish the needs of this experiment was the development of a new feature of Panlab's framework called Federation Computing Interface (FCI) API.

and how Panlab framework is able by means of Federation Computing Interface API to managed resource.

A Use-Case on Testing Adaptive Admission Control 239 2 Use Case Description In order for one to test an adaptive admission control and resource allocation algorithm,

it is necessary to set up an appropriate testbed of a distributed web application like RUBIS benchmark 3,

an auction site prototype modeled after ebay. com. It provides a virtualized distributed application that consists of three components, a web server, an application server, a database and a workload generator,

Furthermore it can be deployed in a virtualized environment using Xen server technology, which allows regulating system resources such as CPU usage and memory,

and provides also a monitoring tool, Ganglia, that measures network metrics, such as round trip time and other statistics,

and resource usage in virtual machines. Fig. 1. The setup for testing the algorithm The adaptive admission control

and resource allocation algorithm is applied to succeed in specific target of network metrics, like round trip time and throughput.

This will be done by deploying a proxy-like control component for admission control and using Xen server technology to regulate CPU usage.

During this scenario the adaptive admission control and resource allocation algorithm is tested against network metrics, like round trip time and throughput.

RUBIS clients will produce requests so that push RUBIS components to their limits, so that resource like CPU usage and network throughput get high values.

During the setup, the researcher wants to test http proxy software written in C programming language that implements an admission algorithm.

Figure 1 displays the 240 C. Tranoris P. Giacomin, and S. Denazis setup for the discussed scenario.

The setup consists of 3 work load http traffic generators, making requests through a hosting unit.

The algorithm, which is located at the proxy unit, needs to monitor the CPU usage of the Web application and Database machines.

Then the algorithm should be able to set new CPU capacity limits on both resources.

Additionally the algorithm should be able to start and stop the work load generators on demand. 3 Technical Environment, Testbed Implementation and Deployment From the requirements of the use case,

it is evident that it would benefit from a testbed offering RUBIS resources. Moreover the experiment needs to manage

and monitor resources within the C algorithm. So the resources need to provide monitoring and provisioning mechanisms.

-Linux machines for the RUBIS based work load generators-A Linux machine for the hosting the algorithm unit,

capable of compiling C and Java software-Linux machines for running XEN server where on top will run the RUBIS Web app

and database The final user needs to provide the algorithm under test. He will just login to the Proxy Unit,

compile the software and execute it. The user will not have access to the RUBIS resources

all the components are managed based on Virtual machines by a XEN server. The implemented RAS instantiate all these Virtual machines and configure the internal components according to end-user needs.

The work load generator exposes parameters such as: used IP for the testbed, memory, hard disk size, number of clients, ramp up time for the requests and a parameter used during the execution of the experiment called Action

which accepts the values start and stop. The Proxy Unit exposes parameters such as used IP for the testbed, memory, hard disk size, username,

password and IP to connect to the RUBIS application resource. The RUBIS application and the RUBIS database have similar parameters to the above and additionally a MON CPU UTILIZATION parameter

which is used to monitor the resource and a CPU CAPACITY used to set the max cpu capacity of the resource.

A Use-Case on Testing Adaptive Admission Control 241 Fig. 2. The Resource adapters of the available testbed resources Fig. 3. RADL definition for the RUBIS application

RADL's textual syntax aims to be easier to describe a RA than code in Java or other target language.

when there is a need to configure a resource that offers an API for configuration. The user can configure the resource through some Configuration Parameters.

A Binding Parameter is a variable that is assigned locally by the resource provider, e g. a local IP ADDRESS.

Figure 3 displays the RADL definition for the RUBIS application server. The Configuration Parameters section describes the exposed parameters to the end user.

Figure 4 displays the use case setup as can be done inside the VCT tool of Panlab.

Three rubis client where selected one rubis proxy, one rubis application and one A Use-Case on Testing Adaptive Admission Control 243 rubis database.

For example the RUBIS clients need to know about the IP of the proxy which hosts the algorithm.

The proxy needs to know the IP of the RUBIS application which also needs a reference to the RUBIS database. 4 Running

and Operating the Experiment The scenario during the experiment utilizes the Federation Computing Interface (FCI) API that Panlab provides 5. Federation Computing Interface (FCI) is an API for accessing resources of the federation.

Fig. 5. Designing the algorithm to operate resources during execution In our testing scenario there is a need to configure resources

or even get monitoring status data properly after the VCT is provisioned and while the testing is in progress.

Figure 5 displays this condition where the System Under Test (SUT) is our algorithm. FCI automatically creates all the necessary code that the end user can then inject inside the algorithm's code.

The end-user needs just to ender his credentials in order 244 C. Tranoris, P Giacomin, and S. Denazis FCI to generate the necessary wrapper classes

An example is given in the following code listing in Java://an example Java federation program public class Main {public static void main (String args){/An example for VCT:

the java listing displays how we can access the resources of this VCT. FCI creates a java class,

called academic07()that we can instantiate in order to get access to the resources. Additionally, for each resource that participates in the VCT java classes are able to provide access.

For example the command myvct. getuop rubis cl 91(.set-ACTION("start";"starts the RUBIS client of the rubis cl 91 resource.

()is able to give back the CPU usage of the database resource. 5 Conclusions The results of running an experiment in Panlab are encouraging in terms of moving the designed algorithms from simulating environments to near production environments.

What is really attractive is that such algorithms can be tested in a best-effort environment with real connectivity issues that cannot be performed easily in simulation environments.

although not comparable currently with similar approaches are really encouraging in terms of moving the designed algorithms from simulating environments to near production environments.

What is really attractive is that such algorithms can be tested in a best-effort environment with real connectivity issues that cannot be performed easily in simulation environments.

The scenario presented can be scaled easily up with many clients and web applications. Also, the proxy under test can be replaced by one or more load balancers.

References 1. Website of Panlab and PII European projects, supported by the European commission in its both framework programmes FP6 (2001-2006) and FP7 (2007-2013:

http://www. panlab. net 2. European commission, FIRE website: Last cited: November 21, 2010, http://cordis. europa. eu/fp7/ict/fire 3. RUBIS, http://rubis. ow2. org/4. RADL, http://trac. panlab

. net/trac/wiki/RADL 5. Federation Computing Interface (FCI), http://trac. panlab. net/trac/wiki/FCI Multipath Routing Slice Experiments in Federated

and Carsten Schmoll1 1 FOKUS-Fraunhofer Institute for Open Communication systems, Berlin, Germany tanja. zseby carsten. schmoll@fokus. fraunhofer. de, 2 University of Wuerzburg

, Institute of Computer science, Wuerzburg, Germany, thomas. zinner christian. schwartz phuoc. trangia@informatik. uni-wuerzburg. de 3 University of Vienna,

The Internet today consist of many heterogeneous infrastructures, owned and maintained by separate and potentially competing administrative authorities.

a) multiple authorities and b) applications with very diverse demands, are likely to stay or even increase in the Internet of the future.

In such an environment federation and virtualization of resources are key features that should be supported in a future Internet.

We believe that this experiment provides a good example use case for the future Internet itself

because we assume that the Internet will consist of multiple different infrastructures that have to be combined in application specific overlays

We also assume that the growing demands will push towards a much better measurement instrumentation of the future Internet.

) Future Internet Assembly, LNCS 6656, pp. 247 258,2011. c The Author (s). This article is published with open access at Springerlink. com. 248 T. Zseby et al. 1

Network Virtualization (NV) techniques 5, 17 allow the establishment of such separate slices on top of a joint physical infrastructure (substrate.

NV enables the parallel and independent operation of application-specific virtual networks (e g. for banking, gaming, web) with their own virtual topology,

and that implement a general data transport service are designated as routing slices 13. Routing slices as an architectural concept is known as Transport Virtualization (TV) 23,24.

These concepts have roots in the work on active networks, where the control plane of a router enabled applications fine-grained control of their own routing 6,

11 and sharing of the resources at the routers using either constant or ad hoc slices 16.

Slices, and routing slices in particular, are made up of shared resources that can be contributed by different administrative authorities.

, a combination of fractions of (virtual) links and (virtual) routers. Due to the fine grained granularity of networking resources,

they permit data transport resource to be accessed without knowledge of their physical or network location.

The control and verification of service level agreements (SLAS) between domains as well as inter-domain security have to be addressed in federated testbeds as well as in the real Internet.

Inter-domain SLA validation would profit from common data formats and data exchange among providers (e g. 8). Intrusion detection systems can increase situation awareness (and with this overall security) by sharing information.

Nevertheless, the operators of the testbeds we considered in our setup are willing to cooperate.

Also the acquisition of specific measurement equipment is often difficult in local labs due to the high costs of such hardware.

and outlook to the enhancements of federated facilities. 2 Experiment Objectives and Requirements for a Concurrent Multipath Transport Alternative multipath transport services in future federated networks might employ concurrent or consecutive

and installation of arbitrary software but is distributed only within Germany, has limited a access, and currently provides no federation method.

Booking of Resources With the SFA software it was possible to book nodes in Planetlab, Planetlab Europe and in the VINI Testbed.

a website dedicated to information about Free Tools for Future Internet Research and Experimentation. The Advanced Network Monitoring Equipment (ANME) deployed by the Onelab project within Planetlab Europe includes precise network cards for active delay measurements using ETOMIC and the continuous monitoring platform (Como

Nevertheless, the achievable accuracy of such measurements depends on the synchronization status of the involved observation points.

and use arbitrary software on the G-Lab nodes. We assume that such features are of interest for many experimenters,

therefore it is not suitable for experiments that require real Internet conditions with regard to scale, delay values,

and hardware to support active and passive high precision measurement. Such an infrastructure helps experimenters to perform measurements

and software tools to the public and to share their experience. Further, free T-Rex seeks to employ standardized instruments to improve the comparability and openness of scientific results in the field of future Internet research.

The platform gives an overview of available tools in future Internet experimental facilities and, based on user feedback,

the tools'feasibility for experiment requirements can be assessed. Another objective is to create links to relevant groups and support standardization efforts in the field of research experiment observation.

Free T-Rex offers such valuable resources like access to the Mome 12 trace and tool database and measurement services, the employed packet tracking service 18, Tophat 9,

we outlined how the federation of multiple experimental facilities can contribute to an improved design of future, federated Internet architectures.

References 1. FIRE-Future Internet Research & Experimentation (2010), Information available at http://ict-fire. eu/2. Free T-REX:

Free Tools for Future Internet Tools and Experimentation (2010), Information available at http://www. free-t-rex. net/3. Onelab-Future Internet Testbeds

Overcoming the internet impasse through virtualization. IEEE Computer, 34 41 (April 2005) 6. Anerousis, N.,Hjlmtysson, G.:

Service level routing on the Internet. In: IEEE GLOBECOM'99, vol. 1, pp. 553 559 (2002) 7. Becke, M.,Dreibholz, T.,Yyengar, J.,Natarajan, P.,Tuexen, M.:

Load Sharing for the Stream Control Transmission Protocol (SCTP), Internet-Draft (2010), http://tools. ietf. org/html/draft-tuexen-tsvwg-sctp-multipath-00

8. Boschi, E.,Denazis, S.,Zseby, T.:A measurement framework for inter-domain sla validation.

Computer networks 36 (1), 21 34 (2001) 12. Mome. Cluster of European Projects aimed at Monitoring and Measurement (2010), Information available at http://www. ist-mome. org/13.

Network virtualization: Breaking the performance barrier. ACM Queue,(Jan./Feb. 2008) 18. Santos, T.,Henke, C.,Schmoll, C.,Zseby.

Let the internet measure itself. ACM SIGCOMM Computer Communication Review 35 (5), 71 74 (2005) 20.

Phuoc Tran-Gia. G-Lab: A Future Generation Internet Research Platform (2008), Information available at http://www. future-internet. eu/21.

Trilogy. Trilogy: Architecting the Future Internet (2010), Information available at http://www. trilogy-project. org/22.

Wischik, D.,Handley, M.,Braun, M. B.:The Resource Pooling Principle. SIGCOMM Comput. Commun. Rev. 38,47 52 (2008), doi:

Re-sequencing Buffer Occupancy of a Concurrent Multipath Transmission Mechanism for Transport System Virtualization. In:

Using Concurrent Multipath Transmission for Transport Virtualization: Analyzing Path Selection. In: Proceedings of the 22nd International Teletraffic Congress (ITC), Amsterdam, Netherlands (Sep. 2010) J. Domingue et al.

) Future Internet Assembly, LNCS 6656, pp. 259 270,2011. The Author (s). This article is published with open access at Springerlink. com. Testing End-to-end Self management in a Wireless Future Internet Environment Apostolos Kousaridas1, George Katsikas1, Nancy Alonistioti1

, Esa Piri2, Marko Palola2, and Jussi Makinen3 1 University of Athens Athens, Greece scan. di. uoa. gr {akousar, katsikas, nancy}@ di. uoa. gr 2 VTT Technical

more diverse and higher performance platform for accomplishing tests and experiments for future Internet new paradigms.

in order to experiment on the improvement of Qos features by using the Self-NET software for self management over a Wimax network environment.

Experimentation, Testing Facilities, Self management, Future Internet, Wimax, Quality of Service 1 Introduction Several network management frameworks have been specified during the last two decades by various standardization bodies

and forums, like IETF, 3gpp, DMTF, ITU, all trying to specify interfaces, protocols and information models by taking into consideration the respective network infrastructure i e.,

, telecom world, the Internet and cellular communications. The current challenge for the network management systems 260 A. Kousaridas et al. is the reduction of human intervention in the fundamental management functions

and the development of the mechanisms that will render the Future Internet network capable of autonomously configuring,

router, access point), is considered potentially as an autonomic element, which is capable of monitoring its network-related state

The effectiveness and the feasibility of various parameters optimization of existing network protocols avoiding manual effort are tested also.

or Monitor-Decide-Execute Cycle (MDE) and consists of the Network Element Cognitive Manager (NECM)

while section 5 concludes this paper. 2 Experimental Facilities Decription The testing facility connecting a fixed Wimax network to the service-aware network is shown in Fig. 1. The Wimax network environment

As regards the Self-NET provision side at Greece Distributed Internet traffic Generator (D-ITG) 8 has been used,

which is a software tool that generates traffic at both Uoa end machines. This is a Java based platform that manipulates two independent entities,

the first is ITGSEND process that undertakes the traffic generation and the latter is ITGRECV process that captures the packets to the receiver.

There are also some contributory entities that assist in improving the traffic simulation by providing log information Testing End-to-end Self management in a Wireless Future Internet Environment 261 Fig. 1. Octopus testbed Wimax

and Self-NET software federation (ITGLOG), printing and plotting specific metrics (ITGDEC, ITGPLOT) and remotely controlling the traffic generation (ITGAPI).

and application layer protocols are supported by this platform such as TCP, UDP, ICMP, DNS, Telnet, and Voip (G. 711, G. 723, G. 729, Voice Activity Detection and Compressed RTP).

The Self-NET project carries out experiments over the Wimax testbed, remotely via the Internet.

The experiment required development of an additional BS control software and deployment of IP routing

We implemented A BS control software (i e. NECM) to allow dynamically collect Wimax link information from the BS

and to control Quality of Service (Qos) settings on the fly. The NECM changes Qos service classes by setting a new configuration to the BS using Simple Network Management Protocol (SNMP.

For example, transmission delay constraints of real-time multimedia streaming are much stricter than that of bulk data transfer.

IEEE 802. 16d 5, the employed Wimax testbed is based on, specifies four different scheduling types, namely Unsolicited Grant Service (UGS), Real-time Polling Service (rtps), Non-real-time Polling Service (nrtps),

BE and nrtps are for delaytolerant data transmission. However, nrtps provides assured bandwidth for the traffic flow

IP ADDRESS, or port number. In our experiments, we used port numbers to classify the IP traffic flows.

and tunneling from and to the Wimax link. Two routers are dedicated on the Octopus testbed for tunneling

and routing IP traffic. The user traffic from the Self-NET experimentation is tunneled by using two IP tunnels over the Internet

and rerouted over the WIMAX air interface at the Octopus testbed. For the test environment provisioning

the IP tunneling (IPIP) and routing was setup at both ends, which requires two routers at the user premises one for sending data to the uplink

and receiving the downlink flows and one for sending to the downlink and receiving from the uplink.

The first tunnel connects the Wimax BS with the Uoa BS Connector (10.1.3.3 10.1.3.1) while the second one connects the Wimax SS with the Uoa SS Connector (10.1.3.4 10.1.3.2),

creating an internal 10.1.3.0/24 network between these network entities. The traffic sent from the Uoa BS Connector (10.1.1.1) is routed over the IPIP tunnel to the Wimax BS Testing End-to-end Self management in a Wireless Future Internet Environment 263 Fig. 3. Network topology

and IPIP tunneling and after the Wireless transmission (DL) to the Wimax SS, the Uoa SS Connector (10.1.2.1) receives the packets via the second tunnel.

The respective procedure occurs for the UL, while the Uoa SS Connector traffic is tunneled to the Wimax SS,

transmitted to the Wimax BS and routed again through IPIP tunnel to the Uoa BS Connector.

During the traffic exchange, the public IPS'are opaque, as the routing procedure explicitly uses the private addresses.

Resource Adapter Description Language (RADL) 9 was used to generate source code for each Resource Adaptor (RA), where, for example,

the Wimax network elements can be considered as available and configurable resources. We decided to use a separate RA for each IP tunneling machine

and service layers cooperation for more efficient end-to-end self management (Fig. 1). The term cooperation is used to describe the collection of the service-level monitoring data and the usage of service-level adaptation actions for efficient network adaptation.

The NECM of the Wimax BS constantly monitors network device statistics (e g.,, UL/DL used capacity, TCP/UDP parameters, service flows),

The Service-level NECM undertakes to collect service-level data. The Service-level NECM could be placed at the service provider's side

The decision making engine of the NDCM filters the collected monitoring data from the network and the service level

In the specific use case the goal of the NDCM Decision making engine is the identification of high average packet error rate (PER) values for the end clients that consume a VOIP service.

Change the priority of 2 k R flows at the Wimax BS. Change the priority of 3 k R flows at the Wimax BS and the codec of 4 k R flows.

Two schemes for the selection of the optimal action have been proposed and they are described below (Fig. 4

and Fig. 5). According to the decision making output the configuration action is transferred either to the Wimax BS NECM

Testing End-to-end Self-Management in a Wireless Future Internet Environment 265 Fig. 4. Decision-making algorithm for configuration action selection Simple Fig. 4 presents

If the PER is lower than a predefined threshold (PER-threshold) the NDCM decides to change all flows from low priority to high priority service class at the WIMAX BS side.

According to the Codec type the NDCM decides the transition to a codec that achieves higher data compression,

thus reducing packet error rate value. If the clients use the less demanding codec, then the change priority solution is checked.

Fig. 5. Decision making algorithm for configuration action selection Advanced 266 A. Kousaridas et al. The above figure (Fig. 5) illustrates the advanced version of the scheme presented above.

The change of the prioritization scheme at the Wimax BS side (e g.,, from low priority to high priority service class.

-G. 711.1: 48 kbps-G. 711.2: 40 kbps-G. 729.3: 8 kbps-G. 729.2: 7 kbps-G. 723.1: 5 kbps

and different combinations of codec types and priorities (high low) have been set in order to measure the arising packet error rate,

However, the Testing End-to-end Self management in a Wireless Future Internet Environment 267 increase rate is not linear

Table 2 presents the reduction of the packet loss rate after the change of the prioritization (from low priority to high priority service class) at the Wimax BS of the 28 Voip flows that use G. 711.1 codec.

G. 711.1 Voip flows that traverse the Wimax BS and face high packet error rate. The modification of the service class prioritization at the BS side (from low priority to high priority class) is not effective,

Testing End-to-end Self management in a Wireless Future Internet Environment 269 Table 6. Qos features improvement after partial (70%)Voip codec change from G. 711.1

packet error rate reduction) that could be achieved by applying the appropriate adaptation considering the network conditions.

Different wireless links and networks have different capabilities and often service implementers and providers do not have a possibility to test their service over various networks of different access technologies.

Our empirical experiments show how a remote wireless link such as Wimax can be used remotely. However, in order to provide a wireless link as a bookable resource for a large set of customers,

An experimental path towards Self management for Future Internet Environments. In: Tselentis, G.,Galis, A.,Gavras, A.,Krco, S.,Lotz, V.,Simperl, E.,Stiller, B. eds.

Towards the Future Internet-Emerging Trends from European Research, pp. 95 104 (2010) 270 A. Kousaridas et al. 3. Website of Panlab and PII European projects, supported by the European commission

Air Interface for Fixed Broadband Wireless Access Systems. IEEE Std. 802.16-2004 (October 2004) 6. Wahle, S.,Magedanz, T.,Gavras, A.:

Towards the Future Internet-Emerging Trends from European Research, pp. 51 62. IOS Press, Amsterdam (2010) 7. Airspan homepage, http://www. airspan. com 8. Distributed Internet traffic Generator, http://www. grid. unina. it/software

/ITG/index. php 9. Resource Adapter Description Language, http://trac. panlab. net/trac/wiki/RADL Part V:

Future Internet Areas: Networks Part V: Future Internet Areas: Networks 273 Introduction Although the current Internet has been extraordinarily successful as a ubiquitous and universal means for communication

and computation, there are still many unsolved problems and challenges some of which have basic aspects. Many of these aspects could not have been foreseen

when the first parts of the Internet were built, but they do need to be addressed now. The very success of the Internet is creating obstacles to the future innovation of both the networking technology that lies at the Internet's core and the services that use it.

In addition, the ossification of the Internet makes the introduction and deployment of new network technologies and services very difficult and very costly.

The aspects which are considered to be fundamentally missing, are: Mobility of networks, services, and devices.

Guaranteeing availability of services according to Service Level Agreements (SLAS) and high-level objectives. Facilities to support Quality of Service (Qos) and Service Level Agreements (SLAS.

Trust Management and Security, privacy and data protection mechanisms of distributed data. An addressing scheme, where identity and location are embedded not in the same address.

Inherent network management functionality, specifically self management functionality. Cost considerations, whereby the overhead of management should be kept under control

since this is a critical part of life-cycle costs. Facilities for the large scale provisioning and deployment of both services and management, with support for higher integration between services and networks.

The content of this book includes three chapters covering some of the above research challenges in Future Internet.

The Challenges for Enhanced Network Self-Manageability in the Scope of Future Internet Development chapter examines perspectives from the inclusion of the autonomicity

and self-manageability features in the scope of Future Internet's (FI) deployment. Apart from the strategic importance for further evolution

Future Internet Areas: Networks management (NM), as FI should possess a considerably enhanced network manageability capability.

but very promising experimental findings, mainly based on the context of a specific use-case for network coverage and capacity optimization, highlighting the way towards developing specific NM-related solutions,

The Efficient Opportunistic Network Creation in the Context of Future Internet chapter is dedicated to the design of Opportunistic Networks.

In the future Internet era mechanisms for extending the coverage of the wireless access infrastructure and service provisioning to locations that cannot be served otherwise

or for engineering traffic whenever the infrastructure network is congested already will be required. Opportunistic Networks are a promising solution towards this direction.

an Architecture for a Sustainable Future Internet chapter describes how to combine optical network technology with Cloud technology

in order to achieve the challenges of Future Internet. The extent of Internet growth and usage raises critical issues associated with its design principles that need to be addressed before it reaches its limits.

Many emerging applications have increasing requirements in terms of bandwidth, Qos and manageability. Moreover, applications such as Cloud computing and 3d-video streaming require optimization

and combined provisioning of different infrastructure resources and services that include both network and IT resources.

As a huge energy consumer, the Internet also needs to have energy-saving functions. Applications critical for society and business or for real-time communication demand a highly reliable, robust,

and secure Internet. Finally, the Future Internet needs to support sustainable business models, in order to drive innovation, competition, and research.

Combining optical network technology with Cloud technology is key to addressing these challenges. In this context, we propose an integrated approach:

Premium advanced networks and IT managed services integrated with the vanilla Internet will ensure a sustainable Future Internet,

Future Internet Areas: Networks 275 The Deployment and Adoption of Future Internet Protocols chapter from the Socioeconomics Area addresses the deployability of network protocols.

The main message of this chapter is that implementation, deployment, and adoption need to be thought about carefully during the design of the protocol,

) Future Internet Assembly, LNCS 6656, pp. 277 292,2011. The Author (s). This article is published with open access at Springerlink. com. Challenges for Enhanced Network Self-Manageability in the Scope of Future Internet Development Ioannis P. Chochliouros1,,

*Anastasia S. Spiliopoulou2, and Nancy Alonistioti3 1 Head of Research Programs Section, Network Strategy and Architecture Dept.,Hellenic Telecommunications Organization S. A. OTE), 99 Kifissias Avenue

, 15124 Maroussi, Athens, Greece ichochliouros@oteresearch. gr 2 Lawyer, General Directorate for Regulatory affairs, Hellenic Telecommunications Organization S. A. OTE), 99 Kifissias

Avenue, 15124 Maroussi, Athens, Greece aspiliopoul@ote. gr 3 Lecturer, National and Kapodistrian University of Athens, Dept. of Informatics and Communications, 15784, Panepistimiopolis, Ilissia, Athens, Greece nancy@di. uoa. gr Abstract.

and self-manageability features in the scope of Future Internet's (FI) deployment. Apart from the strategic importance for further evolution, we also discuss some major future challenges among which is the option for an effective network management (NM),

but very promising experimental findings, mainly based on the context of a specific use-case for network coverage and capacity optimization, highlighting the way towards developing specific NM-related solutions,

Autonomicity, cognitive networks, Future Internet (FI), network manageability, Network Management (NM), self-configuration, self-manageability, self management, situation awareness (SA.

1 Introduction Moving Towards the Future Internet There is an extensive consensus that the Internet, as one of the most critical infrastructures of the 21st century, can critically affect traditional regulatory theories as*Corresponding Author. 278 I. P. Chochliouros, A s. Spiliopoulou,

as the future of the Internet comes into consideration, in parallel with the appearance and/or the development of modern infrastructures,

even greater challenges appear, with many concerns relevant to privacy, security and governance and with a diversity of issues related to Internet's effectiveness and inclusive character.

higher speeds and improved interactivity through the launch of many interactive media-and contentbased applications 2. Nevertheless, such claims necessitate a more secure, reliable, scalable and easily manageable Internet architecture.

If well deployed, the Internet of the future can bring novelty, productivity gains, new markets and growth.

Furthermore, the Internet underpins the whole global economy. The diversity and sheer number of applications and business models supported by the Internet have affected also largely its nature and structure (3

4). The Future Internet (FI) will not be more of the same, but rather appropriate entities incorporating new technologies on a large scale that can unleash novel classes of applications

and related business models 5. If today's Internet is a crucial element of our economy,

FI will play an even more vital role in every conceivable business process. It will become the productivity tool par excellence.

so called Future Internet initiatives around the world working on defining and implementing a new architecture for the Internet intended to overcome existing limitations mostly in the area of networking (6,

7). The complexity of the FI, bringing together large communities of stakeholders and expertise, requires a structured mechanism to avoid fragmentation of efforts

and Internet services 8. The European union (EU) is actually a potential leader in the FI sector 9. Leveraging FI technologies through their use in smart infrastructures offer the opportunity to boost European competitiveness

security and data protection with transparent and democratic governance and control of offered services as guiding principles (10,11). 1. 1 Autonomicity

and Self management Features in Modern Network Design The face of the Internet is continually changing,

The current Internet has been founded on a basic architectural premise that is: a simple network service can be used as a universal means Enhanced Network Self-Manageability in the Scope of Future Internet Development 279 to interconnect intelligent end systems 13.

Thus, it is centred on the network layer being capable of dynamically selecting a path from the originating source of a packet to its ultimate destination, with no guarantees of packet delivery or traffic characteristics.

thus allowing Internet to reach an impressive scale in terms of interconnected devices. However, while the scale has reached not yet its limits

It is now a common belief that current Internet is reaching both its architectural capability

providing a natural complement to the virtualization of resources-by setting up and tearing down composed services, based on negotiated SLAS.

Manageability of the current network typically resides in client stations and servers, which interact with network elements (NES) via protocols such as SNMP (Simple Network Management Protocol).

Furthermore, the diversity of services as well as the underlying hardware and software resources comprise management issues highly challenging, meaning that currently,

a diversity in terms of hardware resources leads to a diversity of management tools (distinguished per vendor).

x) Self-awareness capabilities to support Enhanced Network Self-Manageability in the Scope of Future Internet Development 281 objectives of minimizing system life-cycle costs and energy footprints;(

Furthermore, new wireless sensor network technologies provide options for inclusion of additional intelligence and the capability

Suitable systems with communication and computational capabilities can be integrated into the fabric of the Internet,

providing an accurate reflection of the real world, delivering fine-grained information and enabling almost real-time interaction between the virtual world and real world.

The present Internet model is based on clear separation of concerns between protocol layers, with intelligence moved to the edges,

control and structure communication systems, according to new management schemes and networking techniques without neglecting the advantages of current Internet.

Among the core drivers for the FI are increased reliability, enhanced services, more flexibility, and simplified operation.

The latter calls for including Network Management (NM) issues into the design process for FI principles.

promoted mainly by the necessity of support interoperability between heterogeneous, complex and distributed systems, while it should remain open for further and continuous improvement without the necessity of another disruptive modification in the future.

security, reliability and Enhanced Network Self-Manageability in the Scope of Future Internet Development 283 robustness.

FI design is required to provide answers to a number of current Internet's deficits, especially when the danger of increased complexity is more than evident.

allowing an ever-evolving Internet. Towards realizing this aim, Self-NET considers that a DC-SNM

Enhanced Network Self-Manageability in the Scope of Future Internet Development 285 3 Challenges and Benefits for the Market Sector The implementation-inclusion of suitable cognitive techniques/systems

In many cases, the network operator is obliged to search through vast amounts of monitoring data to find any inconveniences to his network behaviour

That is, by applying self management techniques intending to optimize the network in terms of coverage, capacity, performance etc.

Traffic management configuration of large wireless networks consisting of multiple, distributed NES of varying technologies, is challenging, time 286 I. P. Chochliouros, A s. Spiliopoulou,

In competitive markets, end-users wish to have access to a network offering adequate coverage and services of high quality,

Applying self-aware mechanisms can conduct to network performance optimization in terms of coverage and capacity, optimization of Qos delivered to the end-user and reduction of human intervention 31.

Enhanced Network Self-Manageability in the Scope of Future Internet Development 287 4 Experimental Results for Network Coverage and Optimization In current practice, wireless network planning is a difficult

Several monitoring parameters should be taken into account for optimal coverage and capacity formation, while diverse configuration actions can be available that in many cases are interrelated,

A characteristic use case, particularly studied, was relevant to the challenge for achieving coverage and capacity optimization, for the underlying network.

specific NM problems have been taken into account, under the wider scope of wireless networks coverage and capacity optimization family 32.

In the proposed test-bed, a heterogeneous wireless network environment has been deployed, consisting of several IEEE 802.11 Soekris access points (AP) 33 and an IEEE 802.16 Base Station (BS) 34,

Wi-fi) and multi-RATS (i e. Wifi, Wimax) were located in the corresponding area, consuming a video service delivered by VLC (video LAN client)- based service provider 35.

For the management of the NECMS, a NDCM has been deployed. The cognitive network manager installed per NE has undertaken several distinct actions, that is:(

i) The optimal deployment of a new Wifi AP;(ii) the self-optimization of the network topology through the assisted vertical handover of terminals from loaded to neighbouring-less loaded-APS or BS (s), and;(

Fig. 3. Network Topology of the proposed Use Case for Coverage and Capacity Optimization. Fig. 4 illustrates the total duration of the channel selection that takes place with the activation of an AP. It is shown that Soekris 1 and Soekris 4 need more time for channel selection.

this is Enhanced Network Self-Manageability in the Scope of Future Internet Development 289 0. 000 5. 000 10.000 15.000 20.000 25.000 30.000 35.000 40.000

. 001 0. 001 0. 018 Communication Phase 22.192 1. 711 2. 601 22.405 Monitor Phase 2. 760 2. 561 3

iv) Possibility for automated handover between end-users of heterogeneous wireless technologies (Wifi, Wimax), especially in cases where there is extreme network traffic

new methods (related to embedded and/or autonomous management, virtualization of systems and network resources, advanced and cognitive networking of information objects),

wireless, fixed and IP networks), taking into consideration the next generation Internet environment and the convergence perspective.

The present work has been composed n the context of the Self-NET (Self management of Cognitive Future Internet Elements) European Research Project

Communication on A Public-Private Partnership on the Future Internet. European commission, Brussels (2009) 2. Chochliouros,

The International Journal on Electronic Markets and Business Media 8 (2), 3 8 (1998) 4. Future Internet Assembly (FIA:

Real world Internet (2009), http://rwi. future-internet. eu/index. php/Position paper 5. Afuah, A.,Tucci, C. L.:

Internet Business models and Strategies: Text and Cases. Mcgraw-hill, New york (2000) 6. European Future Internet portal (2010), http://www. future-internet. eu/Enhanced Network Self-Manageability in the Scope of Future Internet

Development 291 7. Blumenthal, M. S.,Clark, D d.:Rethinking the Design of the Internet: The End-to-end Arguments vs. the Brave New world.

ACM Trans. on Internet Techn. 1 (1), 70 109 (2001) 8. Commission of the European communities:

Communication on The Future EU 2020 Strategy. European commission, Brussels (2009) 9. Tselentis, G.,Domingue, L.,Galis, A.,Gavras, A.,et al.:

Towards the Future Internet-A European Research Perspective. IOS Press, Amsterdam (2009) 10. Organization for Economic Co-operation Development (OECD:

The Seoul Declaration for the Future of the Internet Economy. OECD, Paris, France (2008) 11.

Chochliouros, I. P.,Spiliopoulou, A s.:Innovative Horizons for Europe: The New European Telecom Framework for the Development of Modern Electronic Networks and Services.

The Journal of the Communications network (TCN) 2 (4), 53 62 (2003) 12. Commission of the European communities:

Communication on Future Networks and the Internet. European commission, Brussels (2008) 13. Galis, A.,Brunner, M.,Abramowitz, H.:

MANA Position Paper-Management and Service-Aware Networking Architecture (MANA) for Future Internet/Draft 5. 0 (2008) 14.

International Telecommunication Union-Telecommunication Standardization Sector: Rec. M. 3400: TMN Management Functions. ITU-T, Geneva, Switzerland (2000) 15.

Evolution and Structure of the Internet: A Statistical Physics Approach. Cambridge university Press, Cambridge (2004) 16.

Future Generation Internet Architecture (Final Technical Report. The US Air force Research Laboratory (2003) 19. Chochliouros, I. P.,Spiliopoulou, A s.,Georgiadou, E.,Belesioti, M.,et al.:

A Model for Autonomic Network Management in the Scope of the Future Internet. In: Proceedings of the 48th FITCE International Congress, FITCE, Prague, Czech republic, pp. 102 106 (2009) 20.

Future Internet Elements: Cognition and Self management Design Issues. In: Proceedings of the 2nd International Conference on Autonomic Computing and Communication systems, pp. 1 6 (2008) 21.

Raptis, T.,Polychronopoulos, C.,et al.:Technological Enablers of Cognition in Self-Manageable Future Internet Elements.

In: Proceedings of The First International Conference on Advanced Cognitive Technologies and Applications COGNITIVE 2009, pp. 499 504.

Architectural Principles for Synergy of Self management and Future internet Evolutions. In: Proceedings of the ICT Mobile Summit 2009, pp. 1 8. IMC Ltd, Dublin (2009) 23.

Self-NET Project: Deliverable D1. 1: System Deployment Scenarios and Use Cases for Cognitive Management of Future Internet Elements (2008), https://www. ict-selfnet. eu/24.

Agoulmine, N.,Balasubramaniam, S.,Botvitch, D.,Strassner, J.,et al.:Challenges for Autonomic Network Management. In:

Proceedings of IEEE International Conference on Personal Wireless communications, pp. 355 362 (2000) 292 I. P. Chochliouros, A s. Spiliopoulou,

Self management in Future Internet Wireless Networks: Dynamic Resource Allocation and Traffic Routing for Multi-Service Provisioning.

The autonomic computing edge: Can you CHOP UP autonomic computing? IBM Corporation (2008) 30. Prehofer, C.,Bettstetter, C.:

Self-organization in Communication Networks: Principles and Design Paradigms. IEEE Communications Magazine 43 (7), 78 85 (2005) 31.

Mihailovic, A.,Chochliouros, I. P.,Georgiadou, E.,Spiliopoulou, A s.,et al.:Situation Aware Mechanisms for Cognitive Networks.

Proceedings of the International Conference on Ultra Modern Telecommunications (ICUMT-2009), pp. 1 6. IEEE Computer Society Press, Los Alamitos (2009) 32.

AN-100u/UX Single Sector Wireless Access Base Station User Manual (2008) 35. C.:open-source multimedia framework, player and server, http://www. videolan. org/vlc J. Domingue et al.

Eds.):) Future Internet Assembly, LNCS 6656, pp. 293 306,2011. The Author (s). This article is published with open access at Springerlink. com. Efficient Opportunistic Network Creation in the Context of Future Internet Andreas Georgakopoulos, Kostas Tsagkaris, Vera Stavroulaki,

and Panagiotis Demestichas University of Piraeus, Department of Digital Systems, 80, Karaoli and Dimitriou Street, 18534 Piraeus, Greece {andgeorg, ktsagk, veras, pdemest}@ unipi. gr

Abstract. In the future internet era, mechanisms for extending the coverage of the wireless access infrastructure and service provisioning to locations that cannot be served otherwise

or for engineering traffic whenever the infrastructure network is congested already will be required. Opportunistic networks are a promising solution towards this direction.

Opportunistic Networks, Node Selection, Coverage Extension, Capacity Extension, Future Internet. 1 Introduction The emerging wireless world will be part of the Future Internet (FI.

Challenges such as the infrastructure coverage extension or the infrastructure capacity extension, arise. Opportunistic networking seems a promising solution to the problem of coverage extension of the infrastructure

in order to provide service to nodes which normally would be out of the infrastructure coverage or to provide infrastructure decongestion by extending its capacity.

In general, opportunistic networks (ONS) involve nodes and terminals which engage occasional mobility and dynamically configured routing patterns.

this work discusses on the ON creation as a means to provide extended coverage to the infrastructure

and provides the algorithmic solution of the opportunistic node selection problem statement with respect to indicative scenarios such as the opportunistic coverage extension or the opportunistic capacity extension.

in order to evaluate the proposed algorithm and strengthen the proof of concept. Finally, the article concludes with key findings

Efficient Opportunistic Network Creation in the Context of Future Internet 295 Fig. 1. The emerging cognitive wireless world In 5, the selection and navigation of mobile sensor nodes

is investigated by taking into consideration three metrics including coverage, power and distance of each node from a specified area.

In 7, the issue of server selection is being investigated by proposing a node selection algorithm with respect to the worst-case link stress (WLS) criterion.

These works are proposing specific sensor node selection algorithms by taking into consideration attributes such as the area of coverage, the navigation/mobility issues of moving sensors,

For example, mesh networking is used not for the expansion of the coverage of the infrastructure, but for the wireless coverage of an area using various Radio Access Technologies (RATS) 8. Hence,

they are governed not operator. Moreover, ad hoc networking uses peer nodes to form an infrastructure-less, self-organized network 9,

and the application provisioning via the use of various kinds of nodes (e g. cell phones, PDAS,

laptops and other network-enabled devices). Thus, a fitness function is presented which is able to evaluate the eligibility of each candidate node

A prerequisite of each case (e g. opportunistic coverage extension or opportunistic capacity extension) is that the nodes need to have some type of access to the infrastructure,

Efficient Opportunistic Network Creation in the Context of Future Internet 297 3. 2 ON Creation The next phase of the ON lifecycle is the ON creation.

relay nodes (i e. nodes that can be used as routers, even when they do need not to use an application) and the application nodes (i e. the nodes that use a specific application).

Fig. 3 illustrates the opportunistic coverage extension scenario. According to this scenario, a node which acts as a traffic source like a laptop or a camera is out of the coverage of the infrastructure.

As a result, a solution would comprise the creation of an ON in order to serve the out of infrastructure coverage node.

Opportunism primarily lies in the selection for participation in the ON of the appropriate subset of nodes

Efficient Opportunistic Network Creation in the Context of Future Internet 299 Access providers are benefited from the fact that more users can be supported

since congestion situations can be resolved as illustrated in Fig. 4. Fig. 3. Opportunistic coverage extension scenario.

in order to serve the out of infrastructure coverage nodes. To that context in order to gain awareness of the status of the candidate nodes in the vicinity,

The Efficient Opportunistic Network Creation in the Context of Future Internet 301 matrix contains the three factors (i e. energy

a Bluetooth (IEEE 802.15.1) 16 and a high-speed interface (e g. IEEE 802.11 family 17. According to the high-speed interface, each node has a transmission data rate of 15 Mbps. On the other hand,

the Bluetooth interface has a transmission data rate of 1 Mbps but it is used for a rather short-range coverage (e g. 10 meters).

Also, every new message is created at a 30-second interval and has a variable size ranging from 500 to 1500 kilobytes,

depending on the scenario. Messages are created only from specific 3 hosts which are acting as source nodes,

On the other hand, there is a tendency Efficient Opportunistic Network Creation in the Context of Future Internet 303 of significantly lower delivery rates as the message size increases to 1000 and 1500 kilobytes.

and a variable message size ranging from 500 kilobytes to 1000 kilobytes Efficient Opportunistic Network Creation in the Context of Future Internet 305 0 0. 51 1. 52 2

and Future Work This work presents the efficient ON creation in the context of Future Internet.

Operator-governed ONS are a promising solution for the coverage or capacity extension of the infrastructure by providing extra coverage or capacity wherever and whenever needed without the operator having to invest to expensive infrastructure equipment

in order to serve temporary load surge in an area. For the efficient creation of the ON, specific node attributes need to be taken into consideration

References 1. European Telecommunications Standards Institute (ETSI), Reconfigurable Radio Systems (RRS), Summary of feasibility studies and potential standardization topics, TR 102.838

INFOCOM 2006, 25th IEEE International Conference on Computer Communications (2006) 3. Rong, B.,Hafid, A.:

5th International Conference on Mobile Ad hoc and Sensor Networks, Fujian (2009) 4. Bouabdallah, F.,Bouabdallah, N.:

Computer Communications 31,1763 1776 (2008) 5. Verma, A.,Sawant, H.,Tan, J.:Selection and navigation of mobile sensor nodes using a sensor network.

Pervasive and Mobile Computing 2, 65 84 (2006) 6. Chen, H.,Wu, H.,Tzeng, N.:

Optimal node-selection algorithm for parallel download in overlay content-distribution networks. Computer networks 53,1480 1496 (2009) 8. Akyildiz,

I.,Wang, X.,Wang, W.:Wireless mesh networks: a survey. Computer networks 47,445 487 (2005) 9. Akyildiz,

I.,Lee, W.,Chowdhury, K.:CRAHNS: Cognitive radio ad hoc networks. Ad hoc Networks 7, 810 836 (2009) 10.

Computer networks 36,137 151 (2009) 11. Rossberg, M.,Schaefer, G.:A survey on automatic configuration of virtual private networks.

Computer networks (2011) 12. Houidi, I.,Louati, W.,Ameur, W.,Zeghlache, D.:Virtual network provisioning across multiple substrate networks.

Computer networks 55,1011 1023 (2011) 13. Saaty, T. L.:The Analytic Hierarchy Process. Mcgraw-hill, New york (1980) 14.

IEEE 802.11 Wireless Local area networks, http://ieee802. org/11/18. Spyropoulos, T.,Psounis, K.,Raghavendra, C.:Spray and Wait:

An Efficient Routing Scheme for Intermittently Connected Mobile networks. In: ACM SIGCOMM Workshop on Delay-Tolerant Networking, WDTN (2005) Bringing Optical Networks to the Cloud:

An Architecture for a Sustainable Future Internet Pascale Vicat-Blanc1, Sergi Figuerola2, Xiaomin Chen4, Giada Landi5, Eduard Escalona10, Chris Develder3, Anna Tzanakaki6, Yuri

5 Nextworks 6 Athens Information technology 7 SAP Research 8 Poznan Supercomputing and Networking Center 9 INRIA 10 University of Essex 11 Universiteit van

ADVA 14 Alcatel-lucent 15 Telef'onica I+D 16 Telekomunikacja Polska 17 Indian Institute of technology, Bombay Abstract.

Over the years, the Internet has become a central tool for society. The extent of its growth and usage raises critical issues associated with its design principles that need to be addressed before it reaches its limits.

Moreover, applications such as Cloud computing and 3d-video streaming require optimization and combined provisioning of different infrastructure resources and services that include both network and IT resources.

As a huge energy consumer, the Internet also needs to be energyconscious. Applications critical for society and business (e g.,

, health, finance) or for real-time communication demand a highly reliable, robust and secure Internet. Finally, the future Internet needs to support sustainable business models,

in order to drive innovation, competition, and research. Combining optical network technology with Cloud technology is key to addressing the future Internet/Cloud challenges.

In this con-J. Domingue et al. Eds.):) Future Internet Assembly, LNCS 6656, pp. 307 320,2011. c The Author (s). This article is published with open access at Springerlink. com. 308 P. Vicat-Blanc et al. text,

we propose an integrated approach: realizing the convergence of the IT -and optical-network-provisioning models will help bring revenues to all the actors involved in the value chain.

Premium advanced network and IT managed services integrated with the vanilla Internet will ensure a sustainable future Internet/Cloud enabling demanding and ubiquitous applications to coexist.

Future Internet, Virtualization, Dynamic Provisioning, Virtual Infrastructures, Convergence, Iaas, Optical Network, Cloud 1 Introduction Over the years, the Internet has become a central tool for society.

a layered architecture and an agreed upon set of protocols for the sharing and transmission of data over practically any medium.

The Internet's infrastructure is essentially an interconnection of several heterogeneous networks called Autonomous Systems that are interconnected with network equipment called gateways or routers.

Routers are interconnected together through links, which in the core-network segment are mostly based on optical transmission technology,

but also in the access segments gradual migration to optical technologies occurs. The current Internet has become an ubiquitous commodity to provide communication services to the ultimate consumers:

enterprises or home/residential users. The Internet's architecture assumes that routers are stateless and the entire network is neutral.

There is no control over the content and the network resources consumed by each user. It is assumed that users are well-behaving

and business communications as well as general information exchange thanks to emails, the web, Voip, triple play service, etc. the Internet is currently providing a rich environment for social networking and collaboration and for emerging Cloud-based applications such as Amazon's EC2,

Azure, Google apps and others. The Cloud technologies are emerging as a new provisioning model 2. Cloud stands for ondemand access to IT hardware or software resources over the Internet.

Clouds are revolutionizing the IT world 11, but treat the Internet as always available, without constraints and absolutely reliable,

which is yet to be achieved. Analysts predict that in 2020, more than 80%of the IT will be outsourced within the Cloud 9!

With the increase in bandwidth-hungry applications, it is just a matter of time before the Internet's architecture reaches its limits.

The new Internet's architecture should propose solutions for Qos provisioning, management and control, enabling a highly flexible usage of the Internet resources to meet bursty demands.

If the Internet's architecture is redesigned not, not only mission-critical or business applications in the Cloud will suffer,

but even conventional Internet's users will be affected by the uncontrolled traffic or business activity over it.

Bringing Optical Networks to the Cloud 309 Today, it is impossible to throw away what has made the enormous success of the Internet:

the robustness brought by the datagram building block and the end-to-end principle which are of critical importance for all applications.

we propose to improve the current Internet's architecture with the advanced control and management plane that should improve the integration of both new optical transport network technologies

and the virtualization paradigm with dynamic network provisioning as a way towards such a sustainable future Internet.

The proposed architecture for the future Internet will provide a basis for the convergence of networks optical networks in particular with the Clouds while respecting the basic operational principles of today's Internet.

telecom operators have considered methods for dynamic provisioning of high-capacity network-connectivity services tightly bundled with IT resources.

The rest of this chapter exposes the main challenges to be addressed by the future Internet's architecture,

and exploitation of this architecture. 2 Challenges There are various challenges that are driving today's Internet to the limit,

which in turn have to be addressed by the future Internet's architecture. We consider the following six challenges as priorities:

As of today, the users/applications that require bandwidth beyond 1 Gbps are rather common,

with a growing tendency towards applications requiring a 10 Gbps or even 100 Gbps connectivity.

Examples include networked data storage, high-definition (HD) and ultra-HD multimedia-content distribution, large remote instrumentation applications,

But today, these applications cannot use the Internet because of the fair-sharing principle and the basic routing approach.

which is a challenge in today's best-effort Internet. Indeed, IT resources are processing data that should be transferred from the user's premises or from the data repository to the computing resources.

When the Cloud will be adopted largely and the data deluge will fall in it, the communication model offered by the Internet may break the hope for fully-transparent remote access and outsourcing.

The interconnection of IT resources over networks requires well-managed dynamically invoked, consistent services. IT and network should be provisioned in a coordinated way in the future Internet. 3. Deal with the unpredictability and burstiness of traffic:

The increasing popularity of video applications over the Internet causes the traffic to be unpredictable in the networks.

The traffic's bursty nature requires mechanisms to support the dynamic behavior of the services and applications.

Moreover, another important issue is that the popularity of content and applications on the Internet will be more and more sporadic:

the network effect amplifies reactions. Therefore, the future Internet needs to provide mechanisms that facilitate elasticity of resources provisioning with the aim to face sporadic,

seasonal or unpredictable demands. 4. Make the network energy-aware: It is reported in the literature 10,

and this percentage is expected to rapidly grow over the next few years following the growth of the Internet.

Therefore, as a significant contributor to the overall energy consumption of the planet, the Internet needs to be energy-conscious.

The network's service outages and hostile hacks have received significant attention lately due to society's high dependency on information systems.

The current Internet's service paradigm allows service providers to authenticate resources in provider domains

Currently, the business models deployed by telecom operators are focused on selling services on top of their infrastructures. In addition, operators cannot offer dynamic and smooth integration of diversified resources and services (both IT and network) at the provisioning phase.

the proposed architecture introduces the three basic concepts featured by the future Internet: The Virtual Infrastructure concept and its operational model as a fundamental approach to enable the on-demand infrastructure services provisioning with guaranteed performance

This concept has little to do with the way data is processed or transmitted internally, while enabling the creation of containers with associated nonfunctional properties (isolation, performance, protection, etc.).

As stated above, optical network technologies are among the key components for the future Internet.

hence addressing challenge#1. IT resources comprise another important category of future Internet shared resources aggregated in large-scale data centers and providing high computational and storage capacities.

and the service middleware layer. Each layer is responsible for implementing different functionalities covering the full end-to-end service delivery from the service layer to the physical substrate.

Central to this novel architecture is the infrastructure virtualization layer which abstracts, partitions and interconnects infrastructure resources contained in the physical infrastructure layer.

and managing the network resources constituting the Virtual Infrastructure) is closely interacting with the virtualization layer. 3. Finally,

a service middleware layer is introduced to fully decouple the physical infrastructure from the service level.

Network Control Plane NIPS Network+IT Provisioning Services PIP Physical Infrastructure Provider SML Service Middleware Layer VI Virtual Infrastructure VIO Virtual

A company hosts an Enterprise Information system externally on a Cloud rented from a Softwareas-a-Service (Saas) provider.

It also connects heterogeneous data resources in an isolated virtual infrastructure. Furthermore, it supports scaling (up and down) of services and load.

It provides means to continuously monitor what the effect of scaling will be on response time performance, quality of data security, cost aspect, feasibility, etc.

Our architecture will result in a new role for telecom operators that own their infrastructure to offer their optical network integrated with IT infrastructures (either owned by them or by thirdparty providers) as a service to network operators.

, Cloud computing) with complex attributes (e g.,, optimized energy consumption and optimized capacity consumption) and strict bandwidth requirements (e g.,

where distributed computing and storage resources are scaled automatically up and down, with guaranteed high-capacity network connectivity. The enhanced Network Control Plane (NCP+)proposed in our architecture (Fig. 1) offers integrated mechanisms for Network+IT Provisioning Services (NIPS) through the on-demand and seamless provisioning of optical and IT resources.

These procedures are based on a strong inter-cooperation between the NCP+and the service middleware layer (SML) via a serviceto-network interface, named NIPS UNI during the entire VI service life cycle.

These requirements describe not only the characteristics of the required connectivity in terms 19 http://www. ens-lyon. fr/LIP/RESO/Software/vxdl/home. html 316 P. Vicat

+In anycast services the SML provides just a description of the required IT resources (e g. in terms of amount of CPU),

The path computation is performed by dedicated PCES that implements enhanced computation algorithms able to combine both network

Finally, another key element for the control plane is the interaction with the infrastructure-virtualization layer,

In case of inefficiency of the underlying infrastructure, the control plane is able to request the upgrade or downgrade of the virtual resources,

thereby addressing challenge#4. We evaluated an energy efficient routing algorithm (due to space limitations, the detailed algorithm is not 318 P. Vicat-Blanc et al.

Fig. 6. Number of activated fibers. Fig. 7. Number of activated data centers. shown here) from a networked IT use case:

each source site has certain processing demands which need to be satisfied by suitable IT resources (in a data center).

Note that we assume anycast routing, implying that the destination IT resource can be chosen freely among the available ones,

Simulation results (see Fig. 4-6) indicate that our proposed algorithm can decrease the energy consumption by 10%compared to schemes where only IT infrastructure is considered and up to 50%when taking only the network into account

Our goal is to address the six most critical challenges the Internet has to face urgently to support emerging disruptive applications

The overall architectural blueprint complemented by the detailed design of particular components feeds the development activities of the GEYSERS project to achieve the complete software stack

and evaluate prototypes of the different software components creating and managing optical virtual infrastructures. The other goal is to evaluate the performance and functionality of such a virtualized infrastructure in a realistic production context.

Market-Oriented Cloud computing: Vision, Hype, and Reality of Delivering Computing as the 5th Utility. In:

Proceedings of the 2009 9th IEEE/ACM International Symposium on Cluster Computing and the Grid, WASHINGTON DC, USA.

CCGRID'09, p. 1. IEEE Computer Society Press, Los Alamitos (2009), doi: 10.1109/CCGRID. 2009.97 20 Mintotalpower:

minimizing both network -and IT power; Minnetworkpower: only minimizing the network power; Minnetcapacity: minimizing the number of wavelengths needed to establish all requested lightpaths;

only minimizing the energy consumed by the data centers. 320 P. Vicat-Blanc et al. 3. Clark, D.:

The design philosophy of the DARPA internet protocols. SIGCOMM Comput. Commun. Rev. 18,106 114 (1988), doi:

A Novel Architecture for Virtualization and Co-Provisioning of Dynamic Optical Networks and IT Services.

ICT Future Network and Mobile Summit 2011, Santander, Spain (June 2011) 5. Farrel, A.,Vasseur, J. P.,Ash, J.:

A Path Computation Element (PCE)- Based Architecture. RFC 4655 (Informational)( Aug 2006), http://www. ietf. org/rfc/rfc4655. txt 6. Handley, M.:

Why the Internet only just works. BT Technology Journal 24,119 129 (2006), doi: 10.1007/s10550-006-0084-z 7. Koslovski, G.,Vicat-Blanc Primet, P.,Char ao, A s.:

The next Generation Internet, E-business, and E-everything, http://www. aaas. org/spp/rd/ch20. pdf 10.

Future Internet Areas: Services Part VI: Future Internet Areas: Services 323 Introduction The global economy can be characterised under three main sectors.

The primary sector involves transforming natural resources into primary products which then form the raw materials for other industries1.

The economic importance of the service sector is a major motivation for services research both in the software industry and academia.

The Internet of Services is concerned with the creation of a layer within the Future Internet

and technical solutions created under the Internet of Services umbrella. Firstly there is a need to support the needs of businesses in the area.

The Future Internet will be comprised of a large number of heterogeneous components and systems which need to be linked and integrated.

For example, sensor networks will be composed on adhoc collections of devices with low-level interfaces for accessing their status and data online.

Mobile platforms will need to access to external data and functionality in order to meet consumer expectations for rich interactive seamless experiences.

a second driving requirement for the Internet of Services is to provide a uniform conduit between the Future Internet architectural elements through service-based interfaces.

Architectural within a new global communications infrastructure there is a need to determine how a service layer would fit into an overall Future Internet architecture.

Here research focuses on describing services enabling automated 1 http://en. wikipedia. org/wiki/Primary sector of the economy 2 http://en. wikipedia. org/wiki/Secondary sector of the economy 3 http://en

. wikipedia. org/wiki/Tertiary sector of the economy 4 http://www. eurofound. europa. eu/emire/GREECE/TERTIARIZATION-GR. htm 5 http://en. wikipedia. org

Future Internet Areas: Services and semi-automated approaches to service discovery, composition, mediation and invocation.

Cloud computing definitions vary but cloud computing is acknowledged generally to be the provision of IT capabilities, such as computation, data storage and software on-demand, from a shared pool, with minimal interaction or knowledge by users.

Cloud services can be divided into three target audiences: service providers, software developers and users as follows6:

-Infrastructure as a service offering resources such as a virtual machine or storage services. -Platform as a service providing services for software vendors such as a software development platform

or a hosting service. -Software as a service offering applications, such as document processing or email to end-users. Within this section we have three chapters

which cover several of the issues outlined above. The ability to trade IT-services as an economic good is seen as a core feature of the Internet of Services.

In the chapter Butler et al. SLAS Empowering Services in the future Internet the authors discuss this in relation to Service Level Agreements (SLAS.

In particular they claim a requirement for a holistic view of SLAS enabling their management through the whole service lifecycle:

from engineering to decommissioning. An SLA management framework is outlined as a proposal for handling SLAS in the future Internet.

Evidence supporting the claims is provided through experiences in four industrial case studies in the areas of:

which have attracted attention in recent years within the context of the Web. This work has led to the Semantic web,

and extension of the Web which is machine readable. Ontologies and semantics form a part of the next two chapters in this section.

As mentioned above there is an open question on how best to connect the network and service layers in a new communications infrastructure.

Meeting Services and Networks in the future Internet an ontology based approach is taken combined with a simplification of the network layer structure

which replaces several network layers with ontologies providing the foundations for an Autonomic Internet. Linked Data is the Semantic web in its simplest form

and is based on four principles: Use URIS (Uniform Resource Identifiers) as names for things. Use HTTP URIS so that people and machines can look up those names.

using Semantic web standards. Include links to other URIS, so that other resources can be discovered. 6 See http://www. internet-of-services. com/index. php?

id=274&l=0 Part VI: Future Internet Areas: Services 325 Given the growing take-up of Linked Data for sharing information on the Web at large scale there has begun a discussion on the relationship between this technology and the Future Internet.

In particular, the Future Internet Assemblies in Ghent and Budapest both contained sessions on Linked Data.

The final chapter in this section Domingue et al. Fostering a Relationship Between Linked Data and the Internet of Services discusses the relationship between Linked Data and the Internet of Services.

Specifically, the chapter outlines an approach which includes a lightweight ontology and a set of supporting tools.

John Domingue J. Domingue et al. Eds.):) Future Internet Assembly, LNCS 6656, pp. 327 338,2011. The Author (s). This article is published with open access at Springerlink. com. SLAS Empowering Services in the future Internet1 Joe Butler1, Juan Lambea2, Michael Nolan1, Wolfgang Theilmann3, Francesco Torelli4

, Ramin Yahyapour5, Annamaria Chiasera6, and Marco Pistore7 1 Intel, Ireland, {joe. m. butler, michael. nolan}@ intel. com 2 Telefónica Investigación y Desarrollo, Spain, juanlr@tid. com 3 SAP AG, Germany

, wolfgang. theilmann@sap. com 4 ENG, Italy, francesco. torelli@eng. com 5 Technische Universität Dortmund, Germany, ramin-yahyapour@udo. edu

6 GPI, Italy, achiasera@gpi. it 7 FBK, Italy, pistore@fbk. eu Abstract. IT-supported service provisioning has become of major relevance in all industries and domains.

However, the goal of reaching a truly serviceoriented economy would require that IT-based services can be traded flexibly as economic good,

Furthermore, we propose an SLA management framework that can become a core element for managing SLAS in the future Internet.

The service paradigm is a core principle for the Future Internet which supports integration, interrelation and inter-working of its architectural elements.

Besides being the constituting building block of the so-called Internet of Services, the paradigm equally applies to the Internet of things and the underlying technology cloud platform below.

Cloud computing gained significant attention and commercial uptake in many business scenarios. This rapidly growing service-oriented economy has highlighted key challenges

and opportunities in IT-supported service provisioning. With more companies incorporating cloud based IT services as part of 1 The research leading to these results is supported partially by the European community's Seventh Framework Programme (FP7/2001-2013) under grant agreement n°216556.328 J. Butler et al. their own value chain,

e g. the offering of a software service requires infrastructure resources, software licenses or other software services.

We propose an SLA management framework that offers a core element for managing SLAS in the future Internet.

, business, software, and infrastructure) on the other. With a set of four complementary use case studies, we are able to evaluate our approach in a variety of domains

Service Aggregation demonstrates the aggregation of SLA-aware telecommunication and third party web services: how multi-party, multi-domain SLAS for aggregated services can best be offered to customers. egovernment validates the integration of human-based services with those that are based technology,

SLAS Empowering Services in the future Internet 329 The remainder of this paper is organized as follows. Chapter 2 introduces our reference architecture for an SLA management framework.

Chapter 3 discusses the adoption of the framework, within the Future Internet but also in general System Management environments.

) supports arbitrary service types (business, software, infrastructure) and SLA terms, (3) covers the complete SLA and service lifecycle with consistent interlinking of design-time, planning

business, software and infrastructure. The framework communicates to external parties, namely customers who (want to) consume services

On the highest level, we distinguish the Framework Core, Service Managers (infrastructure and software), deployed Service Instances with their Manageability Agents and Monitoring Event Channels.

The Framework Core encapsulates all functionality related to SLA management, business management, and the evaluation of service setups.

Infrastructure-and Software Service Managers contain all service-specific functionality. The deployed Service Instance is the actual service delivered to the customer

and managed by the framework via Manageability Agents. Monitoring Event Channels serve as a flexible communication infrastructure that allows the framework to collect information about the service instance status. Furthermore,

Business SLA Manager Software SLA Manager Infrastructure SLA Manager Business Manager Service Evaluation Infrastructure Service Manager Software Service Manager Customer

3rd Party Manageability Agent Infrastructure Service<<provider relations>><negotiate>><customer relations>>Monitored Event Channel<<control/track>><evaluate>><prepare/manage>><prepare/manage>><publish>><adjust>>Manageability Agent Software Service<<adjust>>deployed infrastructure service deployed software service<<negotiate>>framework core

First we provide a sketch on how the architecture can be applied to the Future Internet.

SLAS Empowering Services in the future Internet 331 3. 1 Adoption Considerations for the Future Internet The SLA management framework architecture can easily be applied to different Future Internet scenarios.

and networking resources, to sensor-like resources in the Internet of things, to services in the Internet of Services,

Assuming to have Manageability Agents for the relevant artefacts in the future Internet, a management environment consisting of SLA

and connected as needed according to the requirements of the involved value chain stakeholders in the respective Future Internet scenario.

In the following use-case chapters we also provide additional configuration examples of the framework. 3. 2 Adoption Considerations for Cloud computing The SLA@SOI framework should become an intrinsic part of each cloud environment,

The ERP hosting use case (Section 4) contains many aspects of a software cloud. 3. 3 Interlinkage with System Management SLA-driven system management is the primary approach discussed in this paper.

and the requested accuracy of the monitoring (significantly impacts on the number of events to be processed).

We assume a virtualisation-enabled data centre style configuration of server capacity, and a broad range of services in terms of relative priority, resource requirement and longevity.

and data service support to other enterprise services and lines of business. This brings varied expectations of availability, mean-time-torecover, Quality of Service, transaction throughput capacity, etc.

run time adjustment decisions on workload migration SLAS Empowering Services in the future Internet 333 for efficiency,

Taking a holistic cost view, it provides fine grained SLA based data to influence future investment decisions based on capital

From an implementation perspective, user interaction is via a web based UI, used by both IT customers and administrators.

Software services could potentially be selected by choosing a virtual machine template which contains pre-loaded applications,

but software layer considerations are considered not core to this Use Case and are dealt more comprehensively with in the ERP Hosting Use Case.

whose role is to carry out the creation of the new virtual machines which constitute the service along with monitoring and reporting for that service.

Such a solution typically consists of a software package (an application) but also some business-level activities,

At the next level, there are the actual software applications, such as for example a hosted ERP SOFTWARE package. At the next level, there are the required middleware components

which are used equally for different applications. At the lowest layer, there are the infrastructure resources, delivered through an internal or external cloud.

Each service layer is associated with a dedicated SLA, containing service level objectives which are specific to this layer.

The Application SLA is mainly about the throughput capacity of the software solution, its response time,

The Middleware SLA specifies the capacity of the middleware components, the response time guarantee of the middleware components

and the costs required for the offering. The Infrastructure SLA specifies the characteristics of the virtual or physical resources (CPU speed, memory,

and storage) and again the costs required for the offering. The use case successfully applies the SLA framework by realizing distinct SLA Managers for the 4 layers and also 4 distinct Service Managers that bridge to the actual support department

the application, the middleware, and the infrastructure artefacts. From a technical perspective, the most difficult piece in the realization of the whole use case was the knowledge discovery about the nonfunctional behaviour of the different components, e g. the performance characteristics of the middleware.

We collected a set of model-driven architecture artefacts, measurements, best practise rules and managed to consistently interlink them

and a demo video are available at 7. SLAS Empowering Services in the future Internet 335 6 Use Case Service Aggregation The main aim of the Service Aggregation use case is the service

-enabling of core Telco services and their addition with services from third parties (as Internet, infrastructure, media or content services).

additionally Service Aggregator integrates software layer (from SLA@SOI framework architecture. And finally Bank prototype is implemented using the top layer, business.

In this way it is necessary to outline also is executed the provision of Telco web service wrappers by Software SLA Manager in an application server

SMS wrappers deployed in the application server of the corresponding virtual machine has to connect and execute different tasks with core mobile network systems that are behind Telefónica Software Delivery Platform (SDP).

The compo 336 J. Butler et al. nents that can be connected also in the use case are the monitors of the services (SMS and Infrastructure services.

To take care about the violations, track interfaces are used to connect the adjustment components in each SLA Manager.

In the new ecosystems of Future internet of services the key will be the exporting and interconnection of services between different parties.

SLA-aware aggregation of telecommunications services introduces a business opportunity for the agile and efficient co-creation of new service offerings and significant competitive advantages to all.

Such a Health & Mobile Service is provided by a so called Citizen Service Center (CSC)

In this context, the SLA between the Government and the CSC regulates the provision of the health, mobile and contact services,

SLAS Empowering Services in the future Internet 337 From the technical point of view, one of the main challenges of this use case has been the modelling of human-provided services,

while typical software/hardware guarantee terms constraint the quality of each single execution of a service, in this use case the guarantee terms constraint the average value of KPIS computed for hundreds of executions

From the evaluation perspective, the application scenario is particularly critical due to sensitive data on the health status of the citizens

and compared with the trends in the real data extracted from the past behaviours of the systems at the service providers.

Further details on this use case are available at 9. 8 Conclusions Service level agreements are a crucial element to support the emerging Future Internet

We explained a generalpurpose SLA management framework that can become a core element for managing SLAS in the future Internet.

and capabilities on arbitrary service artefacts, including infrastructure, network, software, and business artefacts. Four complementary industrial use cases demonstrated the applicability and relevance of the approach.

Use Case research will tackle additional scenarios, especially relevant for the Future Internet. Last, we plan to open up our development activities via an Open source Project.

The first framework version fully published as open source can be found at 5. Open Access.

This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution,

Journal of Internet Engineering 4 (1)( 2010), http://www. jie-online. org/ojs/index. php/jie/issue/view/8 3. Miller, B.:

The autonomic computing edge: Can you CHOP UP autonomic computing? Whitepaper IBM developerworks (March 2008), http://www. ibm. com/developer works/autonomic/library/ac-edge4/4. Theilmann, W.,Winkler, U

.,Happe, J.,Magrans de Abril, I.:Managing on-demand business applications with hierarchical service level agreements.

In: Berre, A j.,Gómez-Pérez, A.,Tutschku, K. eds. FIS 2010. LNCS, vol. 6369, Springer, Heidelberg (2010) 5. SLA@SOI Open source Framework.

First full release by December 2010, http://sourceforge. net/projects/sla-at-soi 6. SLA@SOI project:

egovernment Use Case, http://sla-at-soi. eu/research/focus-areas/use-case-e-government/Meeting Services and Networks in the future Internet Eduardo Santos1

and collaborate with Future Internet researches, like the Autonomic Internet. Keywords: Future Internet, Network Ontology, Post TCP IP, Services Introduction In recent years it has been remarkable the Internet advancement in throughput and the development of different services and application features.

Many of these are supported by the TCP IP protocols architecture, however, the intermediate layers based on the protocols IP, TCP, UDP and SCTP were developed more than 30 years ago,

when the Internet was used just for a limited number of hosts and with a few services support.

Despite the development of the Internet and its wonderful flexibility and adaptability, there were no significant improvements in its Network and Transport layers

resulting in a communication gap between layers 7, 8. Integration of services and networks is an emerging key feature in the future Internet

and there are a lot of studies, proposals and discussions over questions related to a network able of supporting the current and Future Internet communication challenges.

Some of these studies are related to: EFII, FIA, FIRE, FIND, GENI and other groups. Some of these groups are very expressive, for example FIRE,

) Future Internet Assembly, LNCS 6656, pp. 339 350,2011. c The Author (s). This article is published with open access at Springerlink. com. 340 E. Santos et al.

Considering the possibilities for improvements in the current TCP IP architecture with collaboration for the Future Internet, this work is focused in one alternative to the TCP IP protocols, at layers 3 and 4,

and proposals to extend the use of ontology in computer networks to support the communication needs in a better way.

Another aspect that can be placed in the context of the Future Internet is the use of ontology in networks.

Therefore, working with an ontological view over the Network and Transport layers has proved a promising object of research. 1. 1 Ontological Layers Representation The use of ontology at the intermediate layers permits the Internet Application

The Net-Ontology layer has semantic communication, in OWL (Web Ontology Language), with its superior layer and the DL-Ontology layer.

It is responsible to support the Data link communication to guarantee the correct delivery Meeting Services and Networks in the future Internet 341 of data transfer between links.

The main difference between these two layers is that the Net-Ontology layer is responsible to support service needs beyond simple data transfers.

with examples of some Future Internet works that can be integrated with this approach at the intermediate layers.

One application example is the services integration in heterogeneous environment to the devices mobility in 4g networks handovers,

In this scenario, FINLAN approach contributes to the semantic communication between the DOHAND and the DL-Ontology layer, for the handover in 4g networks.

Knowledge management and Virtualisation planes), presented in 3, to better monitor the networks, as the semantic information can directly be handled by the Net-Ontology and DL-Ontology layers.

Using the TCP IP protocols architecture there are some limitations for the software-driven control network infrastructure

These protocols generally just can send information at the data field and do not support semantic in their stacks. 342 E. Santos et al.

can also contribute to the translations of the MBT (Model-based Translator) software package, by the use of the FINLAN formal representation in OWL.

Vissers'position also does sense for some current and Future Internet proposals by the separation of the internal complexities of each layer

This work uses OWL as formal language for this communication as the OWL was adopted by a considerable number of initiatives

One example of the FINLAN ontology use in the future Internet research area is the possibility to support the AUTOI Functional Components communication with (and between) the network elements.

="Thing"/>Subclassof><Subclassof><Class IRI="#Service"/>Class IRI="#Entity"/>Subclassof>Meeting Services and Networks in the future Internet 343 This work shows how FINLAN can contribute with Future Internet researches (using Autoi

as these studies and results are presented in some of our previous works 4 10,16. 2 Contributions to the Future Internet Works The FINLAN project has adherence with some current efforts in the future Internet research area,

Attempting to the alignment with some Future Internet groups proposals, the next section extends possible collaborations that may be implemented in an integrated way with some works.

a service, a content, a network element and even a cloud computing; ID: the unique identifier of each entity;

delivery guarantee, Qos, security and others. 2. 1 Collaboration to the Autoi Planes One of the Autonomic Internet project expectations is to support the needs of virtual infrastructure management to obtain self management

The FINLAN ontology supports the network communication used by the Autoi vcpi (Virtual Component Programming interface) 13,

use of the CPU, memory assignment, packets lost and others. The invocation of the methods can be done by the AMSS,

like virtual routers, can interact between them through the property hasvirtuallink. Collect, Dissemination and Context Information Processing:

and Networks in the future Internet 345 the number of interactions between the context sources and the context clients, diminishing the network effort in some cases.

The Autoi open source implements a scalable and modular architecture to the deployment, control and management of active sessions used by virtual entities.

It consists of one active element and the forwarding engine like a router 15. Its integration with FINLAN can act in some components,

like the Diverter, the Session Broker and the Virtualisation Broker. There are many others but these are essentials.

as maximum and minimum, requisites for an instance (memory size, storage pool size, number of virtual CPUS, 346 E. Santos et al.

Meeting Services and Networks in the future Internet 347 In this proposal, the objects Media, Rules, Behaviour, Relations and Characteristics,

Individual>3 Integration between Services and Networks This section describes how to integrate this project in collaboration with others Future Internet works,

and can not be disregarded to the future of the Internet infrastructure. Autoi modules connections are performed in well defined form using connection handlers or similar classes that uses TCP IP sockets.

Based on the Autoi Java open source, in the ANPI demo, the ANPISDD class is prepared to use the IP and TCP (port 43702) protocols.

as in the following sample code extracted from the ANPISDD. java code. 348 E. Santos et al. public class ANPISDD extends Thread {private Serversocket server;

+""s1=server. accept(;.With the use of the FINLAN library this communication can be done replacing the IP

and DL-ontology layers presented in Fig. 1. This expands the semantic possibilities for the Autoi planes, through the intermediate layers of the networks in the future Internet,

also used to generate the OWL, by the OWL API (version 3. 0. 0. 1451).

and the needs of the data flow that will start. With the understanding of application needs, the Net-Ontology layer, sends to the DL-Ontology layer another OWL object with the requirements of data communication as a way of addressing, for example.

the communication Fig. 4. Overview of FINLAN Library Implementation Meeting Services and Networks in the future Internet 349 is ready to be established,

and the data is sent through the layers also using raw sockets. At the current stage of development the implementation of FINLAN library is made in application level.

Nevertheless, the future intentions are to implement the FINLAN ontology in Linux operating system kernel level,

allowing the facilities in its use in different programming languages, since the methods proposed would be available at the operating system level. 4 Conclusions This paper has presented the FINLAN ontology works in a collaboration perspective with some Future Internet projects.

We have proposed to better meeting of services and networks by approaching services semantically to the network structure.

It was showed how to integrate FINLAN with Future Internet projects, taking Autoi as example, and how the ontological approach can be applied to Future Internet works like monitoring

and content-centric Internet. Future work will implement the FINLAN ontology at the Linux kernel level

and run performance and scalability experiments with different Future Internet projects open implementations. Further work also will do the extension of the scope of the ontological representation,

by modeling the behavior of FINLAN to support requirements in contribution with different Future Internet projects.

We strongly believe that meeting services and networks through the reduction of network layers and

consequently, through the decreasing of users, services and content complexity is a possible way to achieve flexibility in future networks.

Moreover, we expect that ontological approaches can help to build a Future Internet with its real challenges, requirements and new paradigms.

Also to thank the efforts to gather on the state-of-the-art of the Future Internet. Open Access.

Monitoring Service Clouds in the future Internet. In: Towards the Future Internet-Emerging Trends from European Research, p. 115 (2010) 2 FIRE:

FIRE White paper (Aug. 2009), http://www. ict-fireworks. eu/fileadmin/documents/FIRE WHITE PAPER 2009 V3. 1. pdf 350 E. Santos et al. 3 Galis, A.,Denazis

Management Architecture and Systems for Future Internet. In: Towards the Future Internet-A European Research Perspective, p. 112 (2009) 4 Malva, G r.,Dias, E. C.,Oliveira, B c.,Pereira, J. H. S

.,Kofuji, S. T.,Rosa, P. F.:Implementa¸c ao do Protocolo FINLAN. In: 8th International Information and Telecommunication Technologies Symposium (2009) 5 Pereira, F. S f.,Santos, E s.,Pereira, J. H. S.,Rosa, P. F

.,Kofuji, S. T.:Proposal for Hybrid Communication in Local Networks. In: 8th International Information and Telecommunication Technologies Symposium (2009) 6 Pereira, F. S f.,Santos, E s.,Pereira, J. H. S.,Rosa, P. F

.,Kofuji, S. T.:FINLAN Packet Delivery Proposal in a Next Generation Internet. In: IEEE International Conference on Networking and Services, p. 32 (2010) 7 Pereira, J. H. S.,Kofuji, S. T.,Rosa, P. F.:

Distributed systems Ontology. In: IEEE/IFIP New Technologies, Mobility and Security Conference (2009) 8 Pereira, J. H. S.,Kofuji, S. T.,Rosa, P. F.:

Horizontal Address Ontology in Internet Architecture. In: IEEE/IFIP New Technologies, Mobility and Security Conference (2009) 9 Pereira, J. H. S.,Kofuji, S. T.,Rosa, P. F.:

Horizontal Addressing by Title in a Next Generation Internet. In: IEEE International Conference on Networking and Services, p. 7 (2010) 10 Pereira, J. H. S.,Pereira, F. S f.,Santos, E s.,Rosa, P. F

.,Kofuji, S. T.:Horizontal Address by Title in the Internet Architecture. In: 8th International Information and Telecommunication Technologies Symposium (2009) 11 Pereira, J. H. S.,Santos, E s.,Pereira, F. S f.,Rosa, P. F

.,Kofuji, S. T.:Layers Optimization Proposal in a Post-IP Network. International Journal On Advances in Networks and Services, in Press (2011) 12 Rochwerger, B.,Galis, A.,Breitgand, D.,Levy, E.,C'aceres, J.,Llorente,

I.,Wolfsthal, Y.,et al.:Design for Future Internet Service Infrastructures. In: Towards the Future Internet-A European Research Perspective, p. 227 (2009) 13 Rubio-Loyola, J.,Astorga, A.,Serrat, J.,Chai, W. K.,Mamatas, L

.,Galis, A.,Clayman, S.,Cheniour, A.,Lefevre, L.,et al.:Platforms and Software systems for an Autonomic Internet.

In: IEEE Global Communications Conference (2010) 14 Rubio-Loyola, J.,Astorga, A.,Serrat, J.,Lefevre, L.,Cheniour, A.,Muldowney, D.,Davy, S.,Galis

, A.,Mamatas, L.,Clayman, S.,Macedo, D.,et al.:Manageability of Future Internet Virtual Networks from a Practical Viewpoint.

In: Towards the Future Internet-Emerging Trends from European Research, p. 105 (2010) 15 Rubio-Loyola, J.,Serrat, J.,Astorga, A.,Chai, W. K.,Galis

, A.,Clayman, S.,Mamatas, L.,Abid, M.,Koumoutsos, G.:et al.:Autonomic Internet Framework Deliverable D6. 3. Final Results of the Autonomici Approach.

Autoi Project (2010) 16 Santos, E s.,Pereira, F. S f.,Pereira, J. H. S.,Rosa, P. F.,Kofuji, S. T.:

Towards a Content-Centric Internet. In: Towards the Future Internet-Emerging Trends from European Research, p. 227 (2010) J. Domingue et al.

Eds.):) Future Internet Assembly, LNCS 6656, pp. 351 364,2011. The Author (s). This article is published with open access at Springerlink. com. Fostering a Relationship between Linked Data and the Internet of Services John Domingue1, Carlos Pedrinaci1, Maria Maleshkova1, Barry Norton2,

and Reto Krummenacher3 1 Knowledge Media Institute, The Open university, Walton Hall, Milton Keynes, MK6 7aa UK {j. b. domingue, c. pedrinaci, m

. maleshkova}@ open. ac. uk 2 Karlsruhe Institute of technology, Karlsruhe, Germany barry. norton@aifb. uni-karlsruhe. de 3 Semantic Technology Institute, University

We outline a relationship between Linked Data and the Internet of Services which we have been exploring recently.

The Internet of Services provides a mechanism for combining elements of a Future Internet through standardized service interfaces at multiple levels of granularity.

Linked Data is a lightweight mechanism for sharing data at web-scale which we believe can facilitate the management and use of service-based components within global networks.

Keywords: Linked Data, Internet of Services, Linked Services 1 Introduction The Future Internet is a fairly recent EU initiative

which aims to investigate scientific and technical areas related to the design and creation of a new global infrastructure.

An overarching goal of the Future Internet is that the new platform should meet Europe's economic and societal needs.

The Internet of Services is seen as a core component of the Future Internet: The Future Internet is polymorphic infrastructure,

where the boundaries between silo systems are changing and blending and where the emphasis is on the integration, interrelationships and interworking of the architectural elements through new service-based interfaces.

Frederic Gittler, FIA Stockholm The Web of Data is a relatively recent effort derived from research on the Semantic web 1,

whose main objective is to generate a Web exposing and interlinking data previously enclosed within silos.

Like the Semantic web the Web of Data aims to extend the current human-readable Web with data formally represented

so that software agents are able to process and reason with the information in an automatic and 352 J. Domingue et al. flexible way.

This effort however, is based on the simplest form of semantics, RDF (S) 2, and has focused thus far on promoting the publication,

sharing and linking of data on the Web. From a Future Internet perspective a combination of service-orientation and Linked Data provides possibilities for supporting the integration, interrelationship and interworking of Future Internet components in a partially automated fashion through the extensive use of machine

-processable descriptions. From an Internet of Services perspective, Linked Data with its relatively simple formal representations and inbuilt support for easy access and connectivity provides a set of mechanisms supporting interoperability between services.

In fact the integration between services and Linked Data is increasingly gaining interest within industry and academia.

Examples include, for instance, research on linking data from RESTFUL services by Alarcon et al. 3, work on exposing datasets behind Web APIS as Linked Data by Speiser et al. 4,

and Web APIS providing results from the Web of Data like Zemanta1. We see that there are possibilities for Linked Data to provide a common‘glue'as services descriptions are shared amongst the different roles involved in the provision,

aggregation, hosting and brokering of services. In some sense service descriptions as and interlinked with, Linked Data is complementary to SAP's Unified Service Description Language2 5, within their proposed Internet of Services framework3,

as it provides appropriate means for exposing services and their relationships with providers, products and customers in a rich, yet simple manner

which is tailored to its use at Web scale. In this paper we discuss the relationship between Linked Data and services based on our experiences in a number of projects.

Using what we have learnt thus far, at the end of the paper we propose a generalization of Linked Data

and service principles for the Future Internet. 2 Linked Data The Web of Data is based upon four simple principles,

known as the Linked Data principles 6, which are: 1. Use URIS (Uniform Resource Identifiers) as names for things. 2. Use HTTP URIS so that people can look up those names. 3

. When someone looks up a URI, provide useful information, using standards (RDF*,SPARQL). 4. Include links to other URIS,

so that they can discover more things. 1 http://developer. zemanta. com/2 http://www. internet-of-services. com/index. php?

id=288&l=0 3 http://www. internet-of-services. com/index. php? id=260&l=0 Fostering a Relationship between Linked Data and the Internet of Services 353 RDF (Resource Description Framework) is a simple data model for semantically describing resources on the Web.

Binary properties interlink terms forming a directed graph. These terms as well as the properties are described by using URIS.

Since a property can be a URI, it can again be used as a term interlinked to another property.

SPARQL is a query language for RDF data which supports querying diverse data sources, with the results returned in the form of a variable-binding table,

or an RDF graph. Since the Linked Data principles were outlined in 2006, there has been impelled a large uptake most notably by the Linking Open Data project4 supported by the W3c Semantic web Education and Outreach Group.

As of September 2010, the coverage of the domains in the Linked Open Data Cloud is diverse (Figure 1). The cloud now has nearly 25 billion RDF statements

and over 400 million links between data sets that cover media, geography, academia, lifesciences and government data sets.

Fig. 1. Linking Open Data cloud diagram as of September 2010, by Richard Cyganiak and Anja Jentzsch5.

From a government perspective significant impetus to this followed Gordon brown's announcement when he was UK Prime Minister6 on making Government data freely available to citizens through a specific Web of Data portal7 facilitating the creation of a diverse set of citizen-friendly applications. 4 http

://esw. w3. org/Sweoig/Taskforces/Communityprojects/Linkingopendata 5 http://lod-cloud. net/6 http://www. silicon. com/management/public-sector/2010/03

/22/gordon-brown-spends-30mto-plug-britain-into-semantic web-39745620/7 http://data. gov. uk/354 J. Domingue et al.

On the corporate side, the BBC has been making use of RDF descriptions for some time. BBC Backstage8 allows developers to make use of BBC programme data available as RDF.

The BBC also made use of scalable RDF repositories for the back-end of the BBC world cup website9 to facilitate agile modeling 10.

This site was very popular during the event receiving over 2 million queries per day.

Other examples of commercial interest include: the acquisition of Metaweb11 by Google to enhance search,

and the release of the Opengraph12 API by Facebook. Mark Zuckerberg, Facebook's CEO claimed recently that Open Graph was the the most transformative thing we've ever done for the Web 13.3 Services on the Web Currently the world of services on the Web is marked by the formation of two main groups

of services. On the one hand, classical Web services, based on WSDL and SOAP, play a major role in the interoperability within and among enterprises.

Web services provide means for the development of open distributed systems based on decoupled components, by overcoming heterogeneity

and enabling the publishing and consuming of functionalities of existing pieces of software. In particular, WSDL is used to provide structured descriptions for services, operations and endpoints,

while SOAP is used to wrap the XML messages exchanged between the service consumer and provider.

A large number of additional specifications such as WS-Addressing, WS-Messaging and WS-Security complement the stack of technologies.

On the other hand, an increasing number of popular Web and Web 2. 0 applications as offered by Facebook, Google,

Flickr and Twitter offer easy-to-use, publicly available Web APIS, also referred to as RESTFUL services (properly

when conforming to the REST architectural principles 7). RESTFUL services are centred around resources, which are interconnected by hyperlinks

and grouped into collections, whose retrieval and manipulation is enabled through a fixed set of operations commonly implemented by using HTTP.

In contrast to WSDL-based services, Web APIS build upon a light technology stack relying almost entirely on the use of URIS, for both resource identification and interaction,

and HTTP for message transmission. The take up of both kinds of services is hampered, however by the amount of manual effort required

Research on Semantic web services 8 has focused on providing semantic descriptions of services so that tasks such as the discovery, negotiation,

composition and invocation of Web services can have a higher level of automation. These techniques, originally targeted at WSDL services,

and more scalable solutions covering Web APIS as well. 8 http://backstage. bbc. co. uk/9 http://news. bbc. co. uk/sport1/hi/football

/world cup 2010/default. stm 10 http://www. bbc. co. uk/blogs/bbcinternet/2010/07/bbc world cup 2010 dynamic sem. html 11 http://www. freebase. com/12

http://developers. facebook. com/docs/opengraph 13 http://news. cnet. com/8301-13577 3-20003053-36. html Fostering a Relationship between Linked Data

and the Internet of Services 355 4 Linked Services The advent of the Web of Data together with the rise of Web 2 0 technologies and social principles constitute, in our opinion,

lead to a widespread adoption of services on the Web. The vision toward the next wave of services, first introduced in 9 and depicted in Figure 1,

1. Publishing service annotations within the Web of Data, and 2. Creating services for the Web of Data, i e.,

, services that process Linked Data and/or generate Linked Data. We have devoted since then significant effort to refining the vision 10

and implementing diverse aspects of it such as the annotation of services and the publication of services annotations as Linked Data 11,12,

as well as on wrapping, and openly exposing, existing RESTFUL services as native Linked Data producers dubbed Linked Open Services 13,14.

It is worth noting in this respect that these approaches and techniques are different means contributing to the same vision

and the Web of Data through their integration based on the two notions highlighted above. As can be seen in Figure 2 there are three main layers that we consider.

which may be based WSDL or Web APIS, for which we provide in essence a Linked Data-oriented view over existing functionality exposed as services.

Legacy services could in this way be invoked, either Fig. 2. Services and the Web of Data 356 J. Domingue et al. by interpreting their semantic annotations (see Section 4. 1)

or by invoking dedicated wrappers (see Section 4. 2) and RDF information could be obtained on demand.

data from legacy systems, state of the art Web 2. 0 sites, or sensors, which do not directly conform to Linked Data principles can easily be made available as Linked Data.

In the second layer Are linked Service descriptions. These are annotations describing various aspects of the service which may include:

Following Linked Data principles these are given HTTP URIS, are described in terms of lightweight RDFS vocabularies, and are interlinked with existing Web vocabularies.

Note that we have made already our descriptions available in the Linked Data Cloud through iserve these are described in more detail in Section 4. 1. The final layer in Figure 2 concerns services which are able to consume RDF data

(either natively or via lowering mechanisms), carry out the concrete activity they are responsible for, and return the result, if any,

or continue with the activity it is carrying out using these newly obtained RDF triples combined with additional sources of data.

RDF-aware and their functionality may range from RDF-specific manipulation functionality up to highly complex processing beyond data fusion that might even have real-life side-effects.

The use of services as the core abstraction for constructing Linked Data applications is therefore more generally applicable than that of current data integration oriented mashup solutions.

Data-based descriptions of Linked Services allowing them to be published on the Web of Data and using these annotations for better supporting the discovery, composition and invocation of Linked Services.

and SA-REST for Web APIS. To cater for interoperability, MSM represents essentially the intersection of the structural parts of these formalisms.

Additionally, as opposed to most Semantic web services research to date MSM supports both classical WSDL Web services,

as well as a procedural view on the increasing number of Web APIS and RESTFUL services, which appear to be preferred on the Web.

Fostering a Relationship between Linked Data and the Internet of Services 357 Fig. 3. Conceptual model for services used by iserve As it can be seen in Figure 3,

MSM defines Services, which have a number of Operations. Operations in turn have input, output and fault Messagecontent descriptions.

Messagecontent may be composed of mandatory or optional Messageparts. The addition of message parts extends the earlier definition of the MSM as described in 18.

The SAWSDL WSMO-Lite and hrests vocabularies, depicted in Figure 3 with the sawsdl, wl,

hrests extends the MSM with specific attributes for operations to model information particular to Web APIS,

The former is based a web tool that assists users in the creation of semantic annotations of Web APIS,

which are described typically solely through an unstructured HTML Web page. SWEET14 can open any web page and directly insert annotations following the hrests/Microwsmo microformat.

It enables the completion of the following key tasks: 14 http://sweet. kmi. open. ac. uk/358 J. Domingue et al.

which can be republished on the Web. Extraction of RDF service descriptions based on the annotated HTML.

During the annotation both tools make use of the Web of Data as background knowledge so as to identify

since they are adapted to existing sources of Linked Data. The annotation tools are connected both to iserve for one click publication. iserve15,

builds upon lessons learnt from research and development on the Web and on service discovery algorithms to provide a generic semantic service registry able to support advanced discovery over both Web APIS

and WSDL services described using heterogeneous formalisms. iserve is, to the best of our knowledge,

the first system to publish web service descriptions on the Web of Data, as well as the first to provide advanced discovery over Web APIS comparable to that available for WSDL-based services.

Thanks to its simplicity, the MSM captures the essence of services in a way that can support service matchmaking

and invocation and still remains largely compatible with the RDF mapping of WSDL, with WSMOBASED descriptions of Web services, with OWL-S services,

service descriptions are exposed following the Linked Data principles and a range of advanced service analysis and discovery techniques are provided on top.

It is worth noting that as service publication Is linked based on Data principles, application developers can easily discover services able to process

or provide certain types of data, and other Web systems can seamlessly provide additional data about service descriptions in an incremental and distributed manner through the use of Linked Data principles.

One such example Is linked for instance LUF User Feedback) 16, which links service descriptions with users ratings, tags and comments about services in a separate server.

On the basis of these ratings and comments service recommendation facilities have also been implemented17. 15 http://iserve. kmi. open. ac. uk/16 http://soa4all. isoco. net/luf/about/17 http://technologies

. kmi. open. ac. uk/soa4all-studio/consumption-platform/rs4all/Fostering a Relationship between Linked Data and the Internet of Services 359 In summary,

Which Produce and Consume Linked Data In this section we consider the relationship between service interactions and Linked Data;

that is, how Linked Data can facilitate the interaction with a service and how the result can contribute to Linked Data.

In other words, this section is not about annotating service descriptions by means of ontologies and Linked Data,

but about how services should be implemented on top of Linked Data in order to become first class citizens of the quickly growing Linking Open Data Cloud.

Note that we take a purist view of the type of services which we consider.

These services should take RDF as input and the results should be available as RDF;

, service consume Linked Data and service produce Linked Data. Although this could be considered restrictive, one main benefit is that everything is instantaneously available in a machine-readable form.

Within existing work on Semantic web Services, considerable effort is expended often in lifting from a syntactic description to a semantic representation and lowering from a semantic entity to a syntactic form.

Whereas including this information as annotations requires a particular toolset and platform to interpret them, following Linked Data and 18 http://soa4all. isoco. net/spices/about/19 http://technologies. kmi. open. ac. uk/soa4all-studio/360

J. Domingue et al. REST principles allows for re-exposing the wrappers as RESTFUL services so that the only required platform to interact with them is the Web (HTTP) itself.

As a general motivation for our case, we consider the status quo of the services offered over the geonames data set,

a notable and‘lifelong'member of the Linking Open Data Cloud, which are offered primarily using JSON

-and XML-encoded messaging. A simple example is given in Table 1, which depicts an excerpt of a weather report gathered from the station at the Airport in Innsbruck, Austria.

it conveys neither the result's internal semantics nor its interlinkage with existing data sets.

The keys, before each colon, are ambiguous strings that must be understood per API; in Linked Data, on the other hand, geonames itself provides a predicate

and values for country codes and the WGS84 vocabulary is used widely for latitude and longitude information.

(and indeed also within the Ourairports and DBPEDIA Linked Data sets) 20 but the string value does not convey this interlinkage.

A solution more in keeping with the Linked Data principles, as seen in our version of these services,

reusing URIS from Linked Data source for representing features in input and output messages; making explicit the semantic relationship between input and output.

In order to make the statement of this relationship more useful as Linked Data, the approach of Linked Data Services (LIDS) 25 is to URL-encode the input.

For instance, the latitude and longitude and used as query parameters so that the point is represented in a URI forming a new 20 The three identifiers for the Innsbruck Airport resource are http://sws. geonames. org/6299669/,

and http://dbpedia. org/resource/Innsbruck airport, respectively. 21 http://www. linkedopenservices. org/services/geo/geonames/weather/Fostering a Relationship between Linked Data and the Internet of Services

In aligning LOS and LIDS principles, pursued via a Linked Services Wiki22 and a Linked Data and Services mailing list23,

it can first BE POSTED as a new resource (Linked Data and Linked Data Services so far concentrate on resource retrieval

and therefore primarily the HTTP GET verb), in the standard REST style, and then a resource-oriented service can be offered with respect to it.

of describing accepted/expected messages using SPARQL graph patterns. While this is a design decision, it aims at the greatest familiarity and ease for Linked Data developers.

It is not without precedent in semantic service description 26. The authors of 26 use the SPARQL query language to formulate user goals,

and to define the pre-and post-conditions of SAWSDL-based service descriptions, which to some degree, at least conceptually,

Although, the use of SPARQL is similar across different proposals, how the patterns are exploited again offers alternative,

defined again as SPARQL CONSTRUCT queries. Work is ongoing on graph pattern-based discovery and process definition and execution. 22 http://linkedservices. org 23 http://groups google. com/group/linkeddataandservices/24 Currently that the graph patterns contained in this request,

etc. and free of FILTERS. etc. 4 362 J. Domingue et al. 5 Conclusions In this paper we have outlined how Linked Data provides a mechanism for describing services in a machine readable fashion

and enables service descriptions to be connected seamlessly to other Linked Data. We have described also a set of principles for how services should consume

and produce Linked Data in order to become first-class Linked Data citizens. From our work thus far, we see that integrating services with the Web of Data,

as depicted before, will give birth to a services ecosystem on top of Linked Data, whereby developers will be able to collaboratively

and incrementally construct complex systems exploiting the Web of Data by reusing the results of others.

The systematic development of complex applications over Linked Data in a sustainable, efficient, and robust manner shall only be achieved through reuse.

We believe that our approach is a particularly suitable abstraction to carry this out at Web scale.

We also believe that Linked Data principles and our extensions can be generalized to the Internet of Services That is,

to scenarios where services sit within a generic Internet platform rather than on the Web.

These principles are: Global unique naming and addressing scheme-services and resources consumed and produced by services should be subject to a global unique naming and addressing scheme.

This addressing scheme should be easily resolvable such that software clients are able to access easily underlying descriptions.

Linking linking between descriptions should be supported to facilitate the reuse of descriptions and to be able to specify relationships.

Service abstraction building from SOA principles functionality should be encapsulated within services which should have a distinct endpoint available on the Internet, through

which they can be invoked using standard protocols. Machine processability the descriptions of the services and resources should be machine-processable.

Following from the above we believe that the Future Internet will benefit greatly from a coherent approach

which integrates service orientation with the principles underlying Linked Data. We are also hopeful that our approach provides a viable starting point for this.

More generally, we expect to see lightweight semantics appearing throughout the new global communications platform which is emerging through the Future Internet work

and also note that proposals already exist for integrating Linked Data at the network level25.

http://www. soa4all. eu/Fostering a Relationship between Linked Data and the Internet of Services 363 Open Access.

The Semantic web. Scientific American 284 (5), 34 43 (2001) 2. Brickley, D.,Guha, D.,,R. V. eds.:

Linking Data from RESTFUL Services. In: Workshop on Linked Data on the Web at WWW 2010 (2010) 4. Speiser, S.,Harth, A.:

Taking the LIDS off Data Silos. In: 6th International Conference on Semantic Systems (I-SEMANTICS)( October 2010) 5. Cardoso, J.,Barros, A.,May, N.,Kylau, U.:

Towards a Unified Service Description Language for the Internet of Services: Requirements and First Developments.

In: IEEE Int'l Conference on Services Computing, July 2010, pp. 602 609 (2010) 6. Berners-Lee, T.:

Linked Data-Design Issues (July 2006), http://www. w3. org/Designissues/Linkeddata. html 7. Fielding, R. T.:

Architectural styles and the Design of Network-based Software Architectures. Phd Thesis, University of California (2000) 8. Mcilraith, S. A.,Son, T. C.,Zeng, H.:

Semantic web Services. IEEE Intelligent Systems 16 (2), 46 53 (2001) 9. Pedrinaci, C.,Domingue, J.,Krummenacher, R.:

Services and the Web of Data: An Unexploited Symbiosis. In: AAAI Spring Symposium Linked Data Meets Artificial intelligence, March 2010, AAAI Press, Menlo Park (2010) 10.

Pedrinaci, C.,Domingue, J.:Toward The next Wave of Services: Linked Services for the Web of Data.

Journal of Universal Computer science 16 (13), 1694 1719 (2010) 11. Maleshkova, M.,Pedrinaci, C.,Domingue, J.:

Supporting the creation of semantic RESTFUL service descriptions. In: Workshop: Service Matchmaking and Resource Retrieval in the Semantic web at ISWC (November 2009) 12.

Pedrinaci, C.,Liu, D.,Maleshkova, M.,Lambert, D.,Kopecky, J.,Domingue, J.:iserve: a Linked Services Publishing Platform.

In: Workshop: Ontology Repositories and Editors for the Semantic web at ESWC (June 2010) 13. Krummenacher, R.,Norton, B.,Marte, A.:

Towards Linked Open Services and Processes. In: Future Internet Symposium, October 2010, pp. 68 77.14.

Norton, B.,Krummenacher, R.:Consuming Dynamic Linked Data. In: 1st International Workshop on Consuming Linked Data (November 2010) 15.

Benslimane, D.,Dustdar, S.,Sheth, A.:Services Mashups: The New Generation of Web Applications. IEEE Internet Computing 12 (5), 13 15 (2008) 16.

Phuoc, D. L.,Polleres, A.,Hauswirth, M.,Tummarello, G.,Morbidoni, C.:Rapid Prototyping of Semantic Mash-ups Through Semantic web Pipes.

In: 18th Int'l Conference on World wide web, April 2009, pp. 581 590 (2009) 17. Maleshkova, M.,Kopecky, J.,Pedrinaci, C.:

Adapting SAWSDL for Semantic Annotations of RESTFUL Services. In: Workshop: Beyond SAWSDL at OTM, November 2009, pp. 917 926 (2009) 364 J. Domingue et al. 18.

Kopecky, J.,Gomadam, K.,Vitvar, T.:hrests: An HTML Microformat for Describing RESTFUL Web Services.

In: IEEE/WIC/ACM Int'l Conference on Web Intelligence and Intelligent agent Technology, December 2008, pp. 619 625 (2008) 19.

Vitvar, T.,Kopecký, J.,Viskova, J.,Fensel, D.:WSMO-lite annotations for web services. In:

Bechhofer, S.,Hauswirth, M.,Hoffmann, J.,Koubarakis, M. eds. ESWC 2008. LNCS, vol. 5021, pp. 674 689.

Semantic Markup for Web Services. Technical Report, Member Submission, W3c (2004) 21. Fensel, D.,Lausen, H.,Polleres, A.,de Bruijn, J.,Stollberg, M.,Roman, D.,Domingue, J.:

Enabling Semantic web Services-The Web Service Modeling Ontology. Springer, Heidelberg (2006) 22. Farrell, J.,Lausen, H.:

Extended Semantic web Conference (Posters (June 2010) 25. Speiser, S.,Harth, A.:Towards Linked Data Services.

In: Int'l Semantic web Conference (Posters and Demonstrations (November 2010) 26. Iqbal, K.,Sbodio, M. L.,Peristeras, V.,Giuliani, G.:

Semantic Service Discovery using SAWSDL and SPARQL. In: 4th Int'l Conference on Semantics, Knowledge and Grid, December 2008, pp. 205 212 (2008) Part VII:

Future Internet Areas: Content Part VII: Future Internet Areas: Content 367 Introduction One of the major enablers for the evolution to the Future Internet will be the huge volumes of multimedia content.

The new, powerful, low-cost and user friendly capturing devices (e g. mobile phones, digital cameras, IP networked cameras) supported by new multimedia authoring tools will significantly increase the user generated content.

On the other hand, new media sensor networks and tele-immersion applications will further increase the use of automatic generated content.

As a result, the Internet as we know it today will be challenged and a r) evolution towards Media Internet will be initiated.

The Media Internet is defined as the Future Internet variation which supports professional and novice content producers

and is at the crossroads of digital multimedia content and Internet technologies. It encompasses two main aspects:

Media being delivered through Internet networking technologies (including hybrid technologies) and Media being generated, consumed, shared and experienced on the Web.

The Media Internet is evolving to support novel user experiences such as immersive environments including sensorial experiences beyond video

and audio (engaging all the human senses including smell, taste and haptics) that are adaptable to the user, the networks and the provisioned services.

The objective of this section is to offer different views on the processes, techniques and technologies which may pave the way for a Future Media Internet.

First of all, the Future Media Internet should be based on network architectures that can deal with content as a native type

and for this reason the content oriented network architectures for multimedia content delivery will produce a major revolution in the way that content is processed

and delivered though the Internet. One particular case concerns content distributed through hybrid and heterogeneous network architectures,

e g. hybrid broadcast and Internet delivery enhancing the immersive experience of the user beyond the classical digital TV interactivity.

Second, enhancing media encoding technologies is required for the Internet with the objective to maintain the overall integrity,

and adapt the content to the network, delivery device and user, and also optimize the quality of experience over the Internet.

Third one of the areas where high investment in research has taken place in recent years is related to the multimedia and multimodal search and retrieval of multimedia objects over the Internet.

Last but not least, collaborative platforms for the experimentation of socially augmented and mixed reality applications are needed to produce advanced applications for the users,

and social media including personalization and recommendation, is one of the key orientations of future media technologies.

An increasingly large amount of content on the Web, whether multimedia or text is generated collaboratively user content,

of which the quality is not always controllable. In relation to the first point content oriented network architectures, the paper Media Ecosystems:

Future Internet Areas: Content texts, and second to share and deliver his/her own audiovisual content dynamically, seamlessly,

media encoding technologies for the Internet, the objective of the chapter Scalable and Adaptable Media Coding Techniques for Future Internet discusses SVC (Scalable Video Coding)

and MDC (Multiple Description Coding) techniques along with the real experience of the authors of SVC/MDC over P2p networks

and emphasizes their pertinence in Future Media Internet initiatives in order to decipher potential challenges. For the third point

multimodal and multimedia search and retrieval in the future Internet, the chapter Semantic Context Inference in Multimedia Search reviews the latest advances in semantic context inference,

) Future Internet Assembly, LNCS 6656, pp. 369 380,2011. The Author (s). This article is published with open access at Springerlink. com. Media Ecosystems:

and C. Timmerer4 1 CNRS Labri laboratory, University of Bordeaux, France koumaras@ieee. org, daniel. negru@labri. fr 2 Telecommunication Dept.,University

Politehnica of Bucharest (UPB), Romania eugen. borcoci@elcom. pub. ro 3 Institute of Informatics and Telecommunications, NCSR Demokritos, Greece {gardikis, xilouris}@ iit. demokritos

Future Internet, Multimedia Distribution, Content Awareness, Network Awareness, Content/Service Adaptation, Quality of Experience, Quality of Services, Service Composition, Content-Aware Network

and provision of very high-volume video data. Second, the development of advanced networking technologies in the access and core parts,

with Qos assurance is seen. A flexible way of usage based on virtualised overlays can offer a strong support for the transportation of multimedia flows.

Third, the todays'software technologies support the creation and composition of services while being able to take into account information regarding the transport/terminal contexts

Network neutrality has been the foundational principle of the Internet, albeit today is revisited by service providers,

as a mean for quality provision and profit, to allow sustainable new forms of multimedia communications with an increasing importance in the future Internet.

Based on virtualization, the network can offer enhanced transport and adaptation-capable services. This chapter will introduce

and studies are dedicated currently to (re) define the directions which the Future Internet development should follow.

and telecommunication services as described in 3. The strong orientation of user-centric awareness to services

9. The virtualisation as a powerful tool to overcome the Internet ossification by creating overlays is discussed in 10-11.

or multiple core network domains having content aware processing capabilities in terms of Qos, monitoring, media flow adaptation,

They are actually CANENABLED routers, which together with the associated managers and the other elements of the ecosystem, offer content-and context-aware Quality of Service/Experience, adaptation, security,

no duplicates), making it free for (other) data (e g.,, more enhancement layers. The key innovations of this approach to service/content adaptation are distributed,

This is responsible for the actual routers configuration in its own network, based on cooperation with the CAN Manager (CANMGR) belonging to the CAN Provider (CANP.

Management, Control and Data Planes (MPL, CPL, DPL), parallel and cooperating (not represented explicitly in the picture).

The upper data plane interfaces at the CAN layer and transport the packets between the VCAN layer and the Home-Box layer in both directions.

while a more radical approach can also be envisaged towards full virtualization (i e. independent management and control per VCAN).

The media data flows, are classified intelligently at ingress MANES, and associated to the appropriate VCANS

Figure 2 depicts an example to illustrate the principle of Internet parallelization based on VCANS, with focus on the classification process performed at ingress MANES.

Figure 2 shows the process of VCAN negotiation (action 1 on the figure) and installation in the networks (action 2). Then (action 3) MANE1 is instructed how to classify the data packets, based on information as:

the SM@SP instructs the SP/CP servers how to mark the data packets. The information to be used in content aware classification can be:

The data packets are analysed by the classifier, assigned and forwarded to one of the VCANS for further processing.

Special algorithms are needed to reduce the amount of processing of MANE in the data plane based on deep analysis of the first packets of a flow

AS1 AS2 AN HB SP/CP server AS3 VCAN1/MQC1 VCAN2/MQC2 VCAN3/MQC3 L2, L3, L4 headers High level headers

Scalability is achieved by largely avoiding per-flow signalling in the core part of the network. In the new architecture, MANE also can act as content caches,

1) data confidentiality, integrity and authenticity; and 2) intelligent and distributed access control policy-based enforcement.

The evaluation algorithm considers the user flow characteristics CAN policies and present network conditions. In order to attain the required flexibility,

the related security architecture was designed according to the hop-by-hop model 7 on top of the MANES routers.

The second objective will pursue a content-aware approach that will be enforced by MANE routers over data in motion.

and traffic filtering rules by executing security related algorithms over information gathered by the monitoring subsystem.

MANE routers will derive filtering rules from packet inspection and will inform the CANMGR about those computed rules.

Content-aware security technologies typically perform deep content inspection of data traversing a security element placed in a specific point in the network.

The proposed approach differs by being based on MANE routers, which will be used to construct CANS. An example of a traffic filtering rule could be to drop all traffic matching a set composed of:

or co-locating CP's content servers in NPS'premises, nevertheless, an individual CC may also be a private CP.

and ensuring a satisfactory level of Qos for the end users (by appropriating resources to network upgrades etc).

CAN-enabled routers and associated managers offering together content-and context aware Quality of Service/Experience,

User-Centric Future Internet and Telecommunication Services. In: Tselentis, G.,et al. eds.)) Towards the Future Internet, pp. 217 226.

IOS Press, Amsterdam (2009) 4. Schönwälder, J.,et al.:Future Internet=Content+Services+Management. IEEE Communications Magazine 47 (7), 27 33 (2009) 5. Zahariadis, T.,et al.:

Content Adaptation Issues in the future Internet. In: Tselentis, G.,et al. eds.)) Towards the Future Internet, pp. 283 292.

IOS Press, Amsterdam (2009) 6. Huszák, Á.,Imre, S.:Content-aware Interface Selection Method for Multi-Path Video Streaming in Best-effort Networks.

In: Proc. of 16th International Conference on Telecommunications, Marrakech, Morocco, Jul. 2009, pp. 196 201 (2009) 7. Liberal, F.,et al.:

Qoe and*-awareness in the future Internet. In: Tselentis, G.,et al. eds.)) Towards the Future Internet, pp. 293 302.

IOS Press, Amsterdam (2009) 8. Martini, M. G.,et al.:Content Adaptive Network Aware Joint Optimization of Wireless Video Transmission.

IEEE Communications Magazine 45 (1), 84 90 (2007) 9. Baker, N.:Context-Aware Systems and Implications for Future Internet.

In: Tselentis, G.,et al. eds.)) Towards the Future Internet, pp. 335 344. IOS Press, Amsterdam (2009) 10.

Anderson, T.,et al.:Overcoming the Internet Impasse through Virtualization. Computer 38 (4), 34 41 (2005) 11.

Chowdhury, N m.,Boutaba, R.:Network Virtualization: State of the art and Research Challenges. IEEE Communications Magazine 47 (7), 20 26 (2009) 12.

Levis, P.,et al.:The Meta-Qos-Class Concept: a Step Towards Global Qos Interdomain Services.

Proc. IEEE, Softcom, Oct. 2004 (2004) 13. Paris Flegkas, et al. Provisioning for Interdomain Quality of Service:

the MESCAL Approach. IEEE Communications Magazine (June 2005) 14. Timmerer, C.,et al.:Scalable Video Coding in Content-Aware Networks:

Research Challenges and Open Issues. In: Proc. International Tyrrhenian Workshop on Digital communications (ITWDC), Ponza, Italy (September 2010) 15.

ITU-T SG12: Definition of Quality of Experience. TD 109rev2 (PLEN/12), Geneva, Switzerland, 16-25 jan 2007 (2007) J. Domingue et al.

Eds.):) Future Internet Assembly, LNCS 6656, pp. 381 389,2011. The Author (s). This article is published with open access at Springerlink. com. Scalable and Adaptable Media Coding Techniques for Future Internet Naeem Ramzan and Ebroul Izquierdo School of Electronic

Engineering and Computer science, Queen Mary University of London, Mile end, London E1 4ns, United kingdom {Naeem.

Ramzan, Ebroul. Izquierdo}@ elec. qmul. ac. uk Abstract. High quality multimedia contents can distribute in a flexible, efficient and personalized way through dynamic and heterogeneous environments in Future Internet.

Scalable Video Coding (SVC) and Multiple Description Coding (MDC) fulfill these objective thorough P2p distribution techniques.

This chapter discusses the SVC and MDC techniques along with the real experience of the authors of SVC/MDC over P2p networks and emphasizes their pertinence in Future Media Internet initiatives

in order to decipher potential challenges. Keywords: Scalable video coding, multiple description coding, P2p distribution. 1 Introduction Future Media Internet will entail to distribute

and dispense high quality multimedia contents in an efficient, supple and personalized way through dynamic and heterogeneous environments.

Multimedia content over internet are becoming a well-liked application due to users'growing demand of multimedia content and extraordinary growth of network technologies.

A broad assortment of such applications can be found in these days, e g. as video streaming, video conferencing, surveillance, broadcast, e-learning and storage.

In particular for video streaming, over the Internet are becoming popular due to the widespread deployment of broadband access.

In customary video streaming techniques the client-server model and the usage of Content Distribution Networks (CDN) along with IP multicast were the most desirable solutions to support media streaming over internet.

due to a bandwidth bottleneck at the server side from which all users request the content.

In contrast, Peer-to-peer (P2p) media streaming protocols, motivated by the great success of file sharing applications, have attracted a lot of interest in academic and industrial environments.

and end-user characteristics 382 N. Ramzan and E. Izquierdo such as decoding and display capabilities usually tend to be non-homogeneous and dynamic.

and MDC offers an efficient encoding for applications where content needs to be transmitted to many non-homogeneous clients with different decoding and display capabilities.

which is common in best-effort networks such as the Internet, will not interrupt the reproduction of the stream

MDC combined with path/server diversity offers robust video delivery over unreliable networks and/or in peer-to-peer streaming over multiple multicast trees.

The eventual objective of employing SVC/MDC in Future Internet is to maximize the end-users'quality of experience (Qoe) for the delivered multimedia content by selecting an appropriate combination of the temporal, spatial and quality parameters for each client

and MDC source coding techniques in section 2 and 3. Section 4 describes how to adapt SVC for P2p distribution for Future Internet.

3d Scalable and Adaptable Media Coding Techniques for Future Internet 383 wavelet 1 and hybrid video coding 2 techniques.

i e. number of pixels per spatial region in a video frame. Quality scalability, or commonly called SNR (Signal-to-noise-Ratio) scalability,

This is achieved by extraction and decoding of coarsely quantised pixels from the compressed bit-stream.

and offer the basis for spatial and temporal scalability The ST decomposition results in two distinctive types of data:

the resulting data are mapped into the scalable stream in the Scalable and Adaptable Media Coding Techniques for Future Internet 385 bit-stream organisation module,

which creates a layered representation of the compressed data. This layered representation provides the basis for low-complexity adaptation of the compressed bit-steam. 3 Scalable Multiple Description Coding (SMDC) SMDC is a source coding technique,

General principles and different approaches for MDC are reviewed in 5. Approaches for generating multiple descriptions include data partitioning (e g.,

and prioritized in our proposed system 4. 1 Piece Picking Policy The proposed solution is a variation of the"Give-To-Get"algorithm 8,

unless all of them have Fig. 3. Sliding window for scalable video Scalable and Adaptable Media Coding Techniques for Future Internet 387 already been downloaded.

even if the overall download bandwidth is high. This problem is critical if the requested piece belongs to the base layer,

Good neighbours are those peers that own the piece with the highest download rates, which alone could provide the current peer with a transfer rate that is above a certain threshold.

However, every time the window shifts, the current download rates of all the neighbours are evaluated and the peers are sorted in descending order.

1400 1600 0 50 100 150 200 250 300 Time (s) Download Rate (kb/s) Video Download Rate Received Video

Bitrate Fig. 4. Received download rate and received video bitrate for Crew CIF sequence 388 N. Ramzan and E. Izquierdo 5 Multiple Description Coding over P2p Network Most of the work on MDC is proposed for wireless

applications in which there are issues such as hand over of a client to another wireless source is present.

Thus, additional redundancy introduced by using MDC over internet need to be evaluated carefully. Fig. 5. An example of multiple description using scalable video coding A simple way to generate multiple descriptions using scalable video coding is to distribute the enhancement layer NAL units to separate descriptions.

Scalable and Adaptable Media Coding Techniques for Future Internet 389 MDC over SVC is that the receiver/client can make a reproduction of the video

Internet. These coding schemes provide natural robustness and scalability to media streaming over heterogeneous networks.

The amalgamation of SVC/MDC and P2p are likely to accomplish some of the Future Media Internet challenges.

At last, we persuade Future Internet initiatives to take into contemplation these techniques when defining new protocols for ground-breaking services and applications.

Performance evidence of software proposal for Wavelet Video Coding Exploration group, ISO/IEC JTC1/SC29/WG11/MPEG2006/M13146, 76th MPEG Meeting, Montreux

Engineering and Computer science Queen Mary University of London, UK {qianni. zhang, ebroul. izquierdo}@ elec. qmul. ac. uk Abstract.

representation schemes for its semantic context can be constructed by learning from data. In the target representation scheme, metadata is divided into three levels:

a Bayesian network model is built using from a small amount of training data. Semantic inference and reasoning is performed then based on the model to decide the relevance of a video.

and search engines are expected to be able to understand underlying semantics in content and match it to the query.

Such techniques form a key approach to supporting efficient multimedia content management and search in the Internet.

) Future Internet Assembly, LNCS 6656, pp. 391 400,2011. c The Author (s). This article is published with open access at Springerlink. com. 392 Q. Zhang

which could be online databases, a representation scheme for its semantic context is learned directly from data

and will not be restricted to the predefined semantic structures in specific application domains. Another problem hampering bridging the semantic gap is that it is almost impossible to define precise mapping between semantics

While linking low-level features to mid-level concepts are relatively easy to solve using the well-defined algorithms in the state-of-the-art

a Bayesian network model is learned from a small amount of training data. Semantic inference and reasoning is carried then out based on the learned model to decide

and are extracted using algorithms with reasonable performance. The rest of this chapter is organised as follows: Section 2 gives a review on the state-of-the-art techniques on context reasoning for multimedia retrieval task;

Popular techniques related to storing and enforcing high-level information include neural networks, expert systems, statistical association, conditional probability distributions, different kinds of monotonic and nonmonotonic, fuzzy logic, decision trees, static and dynamic

Neural networks are employed in this approach to classify features extracted from video blobs for their classification task.

10,12, in which the object's likelihood can be calculated from the conditional probability of feature vectors.

or content descriptors that can be computed automatically by current machines and algorithms, and the richness,

A subset of the database randomly selected for training purpose is annotated then manually on the high-level query concept.

Fig. 1. Semantic inference work flow One important feature in this module is that the Bayesian network model is constructed automatically using a learning approach based on K2 algorithm 8,

In this algorithm, a Bayesian network is created by starting with an empty network and iteratively adding a directed arc to a given node from each parent node.

This selection criterion is basically a measure of how well the given graph correlates to the data.

Due to the scope of this paper, we give only a brief introduction to K2 algorithm here.

If the reader is interested in more details about this algorithm, please refer to 8. Then in the inference stage,

when an un-annotated data item is present, the Bayesian network model derived from the training stage conducts automatic semantic inferences for the high-level query.

mn)= P (q) n i=1 P (mi q) 5 Experiments The experiments were carried out on the good sized unedited video database.

'On top of that, two high-level queries have been selected carefully considering those commonly exist in the database with reasonable proportions

and this process 398 Q. Zhang and E. Izquierdo took only a few seconds on a PC with Pentium D CPU 3. 40ghz and 2. 00gb of RAM.

TN FP FN 10 2891 4 63 Precision Recall Accuracy ROC area 71.40%13.70%97.74%63.1%Football stadium TP TN FP

FN 4 2950 3 11 Precision Recall Accuracy ROC area 57.10%26.70%99.53%97.1%As it can

the limited accuracy of the mid-level feature extractors and the abstractness and sparse distribution of the query terms throughout the dataset. 6 Conclusions In this chapter an approach for semantic context learning and inference has been presented.

Modelling and inference in this case were carried out using the K2 algorithm. The proposed approach was tested on a large size video dataset.

The obtained results have shown that this approach was capable of extracting very abstract semantic terms that were distributed scarcely in the database.

Beyond pixels: Exploiting camera metadata for photo classification. Pattern recognition 38 (6), 935 946 (2005) 2. Bradshaw, B.:

IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp. 994 999 (1997) 4. Chang, E.,Goh, K.,Sychay, G.,Wu, G.:

Generating semantic visual templates for video databases. In: 2000 IEEE International Conference on Multimedia and Expo, 2000.

Tech. rep.,Institute for Image Data Research, University of Northumbria at Newcastle (1999), http://www. jisc. ac. uk/uploaded documents/jtap-039. doc 8

A bayesian method for the induction of probabilistic networks from data. Machine learning 9 (4), 309 347 (1992) 400 Q. Zhang and E. Izquierdo 9. Fan, J.,Gao, Y.,Luo, H.,Jain, R.:

Mining multilevel image semantics via hierarchical classification. IEEE Transactions on Multimedia 10 (2), 167 187 (2008) 10.

IEEE Computer Society Conference on Computer Vision and Pattern Recognition, vol. 2 (2003) 12. Hoiem, D.,Sukthankar, R.,Schneiderman, H.,Huston, L.:

IEEE Computer Society Conference on Computer Vision and Pattern Recognition, vol. 2 (2004) 13. Kherfi, M. L.,Ziou, D.:

and Signal Processing ICASSP'04, vol. 3, IEEE Computer Society Press, Los Alamitos (2004) 16.

IEEE Transactions on Neural networks 13 (4), 793 810 (2002) 18. Qian, R.,Haering, N.,Sezan, I.:

Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, vol. 1, pp. 200 206 (1999) 19.

IEEE Transactions on pattern analysis and machine intelligence 22 (12), 1349 1380 (2000) 20. Vailaya, A.,Figueiredo, M. A t.,Jain, A k.,Zhang, H. J.:

IEEE Transactions on Image processing 10 (1), 117 130 (2001) 21. Zhu, X.,Wu, X.,Elmagarmid, A k.,Feng, Z.,Wu, L.:

Video data mining: Semantic indexing and event detection from the association perspective. IEEE Transactions on Knowledge and Data engineering, 665 677 (2005) Part VIII:

Future Internet Applications Part VIII: Future Internet Applications 403 Introduction The Future Internet is grounded in the technological infrastructure for advanced networks and applications.

It constitutes a complex and dynamic system and societal phenomenon; it comprises the processes of innovation,

shaping and the actual use of these technologies and infrastructures in private and public organisations, in different sectors of the economy including the service sectors,

and in social networks. Research on the Future Internet therefore includes the development, piloting and validation of high-value applications in domains such as healthcare, energy, transport, utilities, manufacturing and finance.

Increasingly, research and innovation on the Future Internet such as envisaged in the future Internet PPP programme forms part of a diverse, dynamic and increasingly open Future Internet innovation-ecosystem, where different stakeholders such as researchers

businesses, government actors and user communities are brought together to interact and engage in networked and collaborative innovation.

In the field of Future Internet application areas, several research and innovation topics are emerging for the next years.

In particular, there is a need to explore the opportunities provided by Future Internet technologies in various business

Innovation lies at the core of smart enterprises and includes not only products, services and processes but also the organizational model and full set of relations that comprise the enterprise's value network.

The Future Internet should provide enterprises a new set of capabilities, enabling them to innovate through flexibility

Combinations of Future Internet technologies are needed to deliver maximum value and these combinations require the federation and integration of appropriate software building blocks.

A new generation of enterprise systems comprising applications and services are expected to emerge, fine-tuned to the needs of enterprise users by leveraging a basic infrastructure of utility-like software services.

High-value Future Internet applications are also foreseen in the domain of living, healthcare, and energy.

Smart Living is one of the areas where the focus lies clearly on the human user,

and effectiveness of the health value chain e g. through enabling access to and sharing of patient data, secure data exchange between healthcare actors,

The first topic concerns the resources of telecom operators and service providers such as networks, switching, computing and data cen 404 Part VIII:

Future Internet Applications ters which are prominent targets for energy efficiency. The second includes solutions allowing for energy management and reduction of the overall energy consumption.

One of the key developments in this respect is the use of advanced communication and computing infrastructure as part of the Smart Grid.

and facilitated by Internet-based applications and infrastructures based on common platforms. Therefore, cities and urban environments are facing challenges to maintain

when it comes to shaping the demand for advanced Internet-based services. The living labs approach which comprises open

and user driven innovation in large-scale real-life settings opens up a promising opportunity to enrich the experimentally-driven research approach as currently adopted in the future Internet community.

The first chapter Future Internet Enterprise Systems: a Flexible Architectural Approach for Innovation discusses how emerging paradigms,

such as Cloud computing and Software-as-a-service are opening up a significant transformation process for enterprise systems. This transformation arises from commoditization of the traditional enterprise system functions

and is accelerated by new and innovative development methods and architectures of Future Internet Enterprise Systems.

where computational elements referred to as Future Internet Enterprise Resources will directly act and evolve according to what exists in the real world.

The chapter Renewable Energy Provisioning for ICT Services in a Future Internet discusses the Greenstar Network (GSN

GSN is developed to dynamically transport user services to be processed in data centers built in proximity to green energy sources,

and focuses on heavy computing services dedicated to data centers powered completely by green energy, from a large abundant reserve of natural resources in Canada, Europe and the US.

Future Internet Applications 405 The third chapter Smart Cities and Future Internet: towards Cooperation Frameworks for Open Innovation elaborates the concept of smart cities as environments of open

In order to exploit the opportunities of services enabled by the Future Internet for smart cities, there is a need to clarify the way how living lab innovation methods,

user communities and Future Internet experimentation approaches and testbed facilities constitute a common set of resources.

The fourth chapter Smart Cities at the forefront of the Future Internet presents an example of city-scale platform architecture for utilizing innovative Internet of things technologies to enhance the quality of life of citizens.

the infrastructure level, where it provides a common ground for heterogeneous Internet of things facilities that are interworking;

and at the service level, where the platform can be used to interconnect with different Internet of Services testbeds,

) Future Internet Assembly, LNCS 6656, pp. 407 418,2011. The Author (s). This article is published with open access at Springerlink. com. Future Internet Enterprise Systems:

A Flexible Architectural Approach for Innovation Daniela Angelucci, Michele Missikoff, and Francesco Taglino Istituto di Analisi dei Sistemi ed Informatica A. Ruberti Viale Manzoni 30,

In recent years, the evolution of infrastructures and technologies carried out by emerging paradigms, such as Cloud computing,

Future Internet and Saas (Software-as-a-service), is leading the area of enterprise systems to a progressive, significant transformation process.

This process will be accelerated by the advent of FINES (Future Internet Enterprise System) research initiatives, where different scientific disciplines converge,

according to the different articulations that Future Internet Systems (FIS) are assuming, to achieve the Future Internet Enterprise Systems (FINES).

In particular, this paper foresees a progressive implementation of a rich, complex, articulated digital world that reflects the real business world,

where computational elements, referred to as FINER (Future Internet Enterprise Resources), will directly act and evolve according to what exists in the real world.

Future Internet, Future Enterprise Systems, component-based software engineering, COTS, SOA, MAS, smart objects, FINES, FINER. 1 Introduction In recent years, software

development methods and technologies have evolved markedly, with the advent of SOA 15, MDA 16, Ontologies and Semantic web,

and F. Taglino (Future Internet Enterprise Systems) Research Roadmap1, a study carried out in the context of the European commission,

Internet of things and Enterprise Environments (DG Infso. The report claims that we are close to a significant transformation in the enterprise systems, where

This movement is facilitated further by the evolution of infrastructures and technologies, starting from Cloud computing and Future Internet,

and realising enterprises software applications. In essence, while enterprise management and planning services will be increasingly available from the‘cloud',in a commoditised form,

Future Internet, Web 2. 0, Semantic web, Cloud computing, Saas, Social media, and similar emerging forms of distributed, open computing will push forward new forms of innovation such as,

and in particular, Open Innovation 3. The quest for continuous, systematic business innovation requires (I ES capable of shifting the focus to ideas generation and innovation support,

New business requirements that current software engineering practices do not seem to meet. Therefore we need to orientate the research towards new ES architectures and development paradigms

To this end, the ICT domain needs to push forward the implementation of future ES development environments,

Such development environments will be based on an evolution of MDA, being able to separate the specification and development of the (i) strategic business logic from the (ii) specific business operations and,

and advanced graphical user 1 http://cordis. europa. eu/fp7/ict/enet/documents/task-forces/research-roadmap/Future Internet Enterprise Systems 409

interfaces will foster new development environments conceived for business experts to directly intervene in the development process.

The second grand research challenge concerns the architecture of the Future Internet Enterprise Systems (FINES) that need to deeply change with respect to

pushed by the new solutions offered in the future Internet Systems (FIS) field. In particular, we may mention:

the Internet of Services (Ios), Internet of things (Iot) and smart objects, Internet of Knowledge (Iok), Internet of People (Iop.

(and outside) an enterprise will have a digital image (a sort of‘avatar')that has been referred to as Future Internet Enterprise Resource (FINER) in the FINES Research Roadmap.

Together, they need to cooperate in developing a new breed of services, tools, software packages, interfaces and user interaction solutions that are not available at the present time.

In particular on the first and the second GRC that concern the development of new FINESS capable of offering to the business experts the possibility of directly governing the development of software architectures.

if such software architectures will correspond to the enterprise architectures, and will be composed by elements tightly coupled with business entities.

from Cloud computing to Social media, to Service-oriented Computing, from Business Process Engineering to semantic technologies and mash-up.

seen as the new frontier to software components aimed at achieving agile system architectures. Section V provides some conclusions

Traditionally, the software engineering community has devoted great attention to design approaches, methods and tools, supporting the idea that large software systems can be created starting from independent,

reusable collections of preexisting software components. This technical area is referred often to as Component Based Software engineering (CBSE.

The basic idea of software componentization is quite the same as software modularization, but mainly focused on reuse.

CBSE distinguishes the process of"component development"from that of"system development with components 9. CBSE laid the groundwork for the Object oriented Programming (OOP) paradigm that in a short time imposed itself over the preexisting modular software development techniques.

OOP aims at developing applications and software systems that provide a high level of data abstraction and modularity (using technologies such as COM,.

, NET, EJB and J2ee. Another approach to componentization is that of the Multi Agent Systems (MAS),

which is based on the development of autonomous, heterogeneous, interacting software agents. Agents mark a fundamental difference from conventional software modules in that they are inherently autonomous and endowed with advanced communication capability 10.

On the other side, the spread of the Internet technologies and the rising of new communication paradigms, has encouraged the development of loosely coupled and highly interoperable software architectures through the spread of the Service-Oriented approach,

and the consequent proliferation of Service-Oriented Architectures (SOA). SOA is an architectural approach whose goal is to achieve loose coupling among interacting software services, i e.,

, units of work performed by software applications, typically communicating over the Internet 11. In general, a SOA will be implemented starting from a collection of components (e-services) of two different sorts.

Some services will have a‘technical'nature, conceived to the specific needs of ICT people; some other will have a‘business'nature,

reflecting the needs of the enterprise. Furthermore, the very same notion of an eservice is an abstraction that often hides the entity

(or agent) that in the real world provides such a service. Such an issue may seem trivial to ICT people (they need a given computation to take place;

where it is performed or who is taking care of it is inconsequential). Conversely, for business people, services are generated not‘in the air':

'there is an active entity (a person, an organization, a computer, a robot, etc. that provides the services, with a given cost and time (not to mention SLA, etc.

Future Internet Enterprise Systems 411 In summary, Web services were introduced essentially as a computation resource,

Our aim to achieve an agile system architecture made up of FINERS put its basis upon the spread of the Cloud computing philosophy

where business expert can directly manage a new generation enterprise software architectures. Cloud computing represents an innovative way to architect

and remotely manage computing resources: this approach aims at delivering scalable IT resources over the Internet,

as opposed to hosting and operating those resources (i e. applications, services and the infrastructure on

which they operate) locally. It refers to both the applications delivered as services over the Internet

and the hardware and system software in the datacenters that provide those services 12. Cloud computing may be considered the basic support for a brand new business reality where FINERS can easily be searched,

composed and executed by a business expert. FINERS will implement a cloud-oriented way of designing,

organizing and implementing the enterprises of the future. In conclusion for decades component technologies have been developed with an ICT approach,

to ease software development processes. Conversely, we propose to base a FINES architecture on building blocks based on business components.

net Future Internet Enterprise Systems 413 worked structure, conceived as an evolution of the Linked Open Data2 of today;

, according to IPV6, URI3, or ENS4. GR: Graphical Representation. This can vary from a simple GIF to a 3d model,

with the protocols for issuing (as client) or responding (as server) to request messages. It is structured according to the grounding of OWL-S. 2 http://esw. w3. org/Sweoig/Taskforces/Communityprojects/Linkingopendata 3 Universal Resource Identifier,

Tangible entity, from computers to aircrafts, to buildings and furniture. Intangible entity, for which a digital image is mandatory.

Fig. 2. The FINER Pentagone All these FINERS will freely interact and cooperate, according to what happens for their real world counterparts.

and will be connected constantly (transparently, in a wired or wireless mode) to the Internet, to reach other FINERS,

Future Internet Enterprise Systems 415 5. 1 A Business-Driven FINES Develpment Platform In order to put the business experts at the centre of the ES development process, we foresee a platform

and are reached through the Internet. On the FINES development environment (see Fig. 3), FINERS are represented visually in a 3d space that models the enterprise reality (i e.,

a Virtual Enteprise Reality) where the user can navigate and manage changes. At a lower level, simpler FINERS will be aggregated to form more complex ones.

Future Internet will play a central role in supporting the discovery of the needed FINERS that often will be acquired virtually (in case of intangible assets),

and data generated during FINERS'operations. There is not a centralised database, the information will stay by the business entity to

which they pertain or in the Cloud. A similar interface, representing a Virtual Enterprise Reality

The computational resources of a FINES are maintained in the Computing Cloud, and are linked recursively to compose complex FINERS starting from simpler ones.

FINERS Cloud Space Real world Low Level FINERS EVENT RESPONSE High Level FINERS Fig. 4. FINES Runtime Environment Future Internet Enterprise

summarised in the sentence‘The Network is the Computer'.'As it happens with early intuitions,

As a next prophecy we propose the Enterprise is the Computer, meaning that an enterprise,

and operational, will enjoy a fully distributed computing power, where computation will be performed directly by enterprise components,

mainly positioned in the enterprise itself of in the Cloud (typically, in case of intangible entities).

and maintaining large scale computing solutions simply interacting with a familiar (though technologically enhanced) business reality.

Iot, Ios, Multi-Agent Systems, Cloud computing, Autonomic Systems) and, in parallel, some key areas of the enterprise that will start to benefit of the FINES approach.

Strategies and Challenges for Interconnecting Wireless Mesh and Wireless Sensor Networks. Wireless Personal Communications 53 (3)( 2010) 2. Buxmann, P.,Hess, T.,Ruggaber, R.:

-Internet of Services. Business & Information systems Engineering 1 (5), 341 342 (2009) 3. Chesbrough, H.:

Open Innovation: The new Imperative for Creating and Profiting from Technology. Harvard Business school Press (2003) 418 D. Angelucci, M. Missikoff,

Proceedings of the 2008 international workshop on Software engineering for adaptive and self-managing systems (2008) 8. Villa, F.,Athanasiadis,

Environmental Modelling & Software 24 (5)( 2009) 9. Crnkovic, I.,Larsson, S.,Chaudron, M.:Component-based Development Process and Component Lifecycle.

27th International Conference on Information technology Interfaces (ITI), Cavtat, Croatia, IEEE, Los Alamitos (2005) 10. Nierstrasz, O.,Gibbs, S.,Tsichritzis, D.:

Component-oriented software development, Special issue on alaysis and modeling in software development, pp. 160 165 (1992) 11.

A Berkley View of Cloud computing, EECS-2009-28 (2009) 13. Martin, D.,et al.:Bringing Semantics to Web Services with OWL-S. In:

Proc. Of WWW Conference (2007) 14. Clark, D.,et al.:Newarch project: Future-generation internet architecture.

Tech Rep. MIT Laboratory for Computer science (2003), http://www. isi. edu/newarch/15. Tselentis, G.,et al.

eds.):) Towards the Future Internet-Emerging Trends from European Research. IOS Press, Amsterdam (2010) 16.

Papazoglou, M. P.:Web Services: Principles and Technology. Prentice-hall, Englewood Cliffs (2007) 17. Mellor, S. J.,Scott, K.,Uhl, A.,Weise, D.:

Model-driven architecture. In: Bruel, J.-M.,Bellahsène, Z.,et al. eds.)) OOIS 2002. LNCS, vol. 2426, p. 290.

Springer, Heidelberg (2002) J. Domingue et al. Eds.):) Future Internet Assembly, LNCS 6656, pp. 419 429,2011.

The Author (s). This article is published with open access at Springerlink. com. Renewable Energy Provisioning for ICT Services in a Future Internet Kim Khoa Nguyen1, Mohamed Cheriet1, Mathieu

Lemay2, Bill St. Arnaud3, Victor Reijs4, Andrew Mackarel4, Pau Minoves5, Alin Pastrama6, and Ward Van Heddeghem7 1 Ecole de Technologie Superieure, University of Quebec, Canada kim. nguyen@synchromedia. ca, Mohamed.

(GSN) is developed to dynamically transport user services to be processed in data centers built in proximity to green energy sources, reducing GHG (Greenhouse Gas) emissions of ICT equipments.

the heaviest computing services are dedicated to virtual data centers powered completely by green energy from a large abundant reserve of natural resources,

Green Star Network, Mantychore FP7, green ICT, Future Internet 1 Introduction Nowadays, reducing greenhouse gas (GHG) emissions is becoming one of the most challenging research

Research projects following this direction have focused on microprocessor design, computer design, power-on-demand architectures and virtual machine consolidation techniques.

Large ICT companies, like Microsoft which consumes up to 27mw of energy at any given time 1,

have built their data centers near green power sources. Unfortunately many computing centers are not so close to green energy sources.

Thus, green energy distributed network is an emerging technology, given that losses incurred in energy transmission over power utility infrastructures are much higher than those caused by data transmission,

which makes relocating a data center near a renewable energy source a more efficient solution than trying to bring the energy to an existing location.

The Greenstar Network (GSN) project 3 is one of the first worldwide initiatives aimed at providing ICT services based entirely on renewable energy sources such as solar wind and hydroelectricity across Canada and around the world.

The network can transport user service applications to be processed in data centers built in proximity to green energy sources

, such as hand-held devices, home PCS), the heaviest computing services will be dedicated to data centers powered completely by green energy.

and in the server farms is considered not since no special equipment is deployed in the GSN.

In order to move virtualized data centers towards network nodes powered by green energy sources distributed in such a multi-domain network, particularly between Europe and North america domains,

Management and technical policies will be developed to leverage virtualization which helps to migrate virtual infrastructure resources from one site to another based on power availability.

This will facilitate use of renewable energy within the GSN providing an Infrastructure as a service (Iaas) management tool.

Services in a Future Internet 421 one is powered by a different renewable energy source) could be integrated into an everyday network.

Core nodes are linked by an underlying high speed optical network having up to 1 000 Gbit/s bandwidth capacity provided by CANARIE.

in comparison to electronic equipments such as routers and aggregators 4. The migration of virtual data centers over network nodes is indeed a result of a convergence of server and network virtualizations as virtual infrastructure management.

During the service, the user monitors and controls resources as if he was the owner, allowing the user to run their application in a virtual infrastructure powered by green energy sources. 2 Provisioning of ICT Services over Mantychore FP7

and MANTICORE II 8 9. The initial MANTICORE project goal was to implement a proof of concept based on the idea that routers

MANTICORE II continued in the steps of its predecessor to implement stable and robust software while running trials on a range of network equipment.

which provides complete control of optical resources. b) Layer 2, Ethernet and MPLS. Users will be able to get control over Ethernet

and MPLS (Layer 2. 5) switches to configure different services. In this aspect, Mantychore will integrate the Ether project 6 and its capabilities for the management of Ethernet and MPLS resources. c) Layer 3

Mantychore FP7 suite includes set of features for: i) Configuration and creation of virtual networks, ii) Configuration of physical interfaces, iii) Support of routing protocols, both internal (RIP, OSPF) and external (BGP), iv) Support of Qos

and firewall services, v) Creation, modification and deletion of resources (interfaces, routers) both physical and logical,

and vi) Support of IPV6. It allows the configuration of IPV6 in interfaces, routing protocols, networks.

Fig. 1. The Greenstar Nework Figure 1 shows the connection plan of the GSN. The Canadian section of the GSN has the largest deployment of six GSN nodes powered by sun, wind and hydroelectricity.

Building competency Renewable Energy Provisioning for ICT Services in a Future Internet 423 using renewable energy resources is vital for any NREN with such an abundance of natural power generation

houses a GSN Node at a data center in Reykjavík (Iceland) and also contributes to the GSN/Mantychore controller interface development.

The only difference between the GSN and a regular network is that the former one is able to transport ICT services to data centers powered Switch (Allied Telesis) Raritan UPS (APC) Gbe Tranceiver

PDU Servers (Dell Poweredge R710) To core network Wind power node architecture (Spoke) Switch (Allied Telesis) Raritan UPS (APC) PDU Servers (Dell

Poweredge R710) Hydroelectricity power node architecture (Hub) MUX/DEMUX To core network Backup Disk Arrays Gbe Tranceiver MUX/DEMUX GSN-Montreal

wind and solar types) 424 K. K. Nguyen et al. by green energy and adjust the network to the needs controlled by software.

such as routers and servers, is considered not, because no special hardware equipment is used in the GSN.

Figure 2 illustrates the architectures of a hydroelectricity and two green nodes one is powered by solar energy

The solar panels are grouped in bundles of 9 or 10 panels, each panel generates a power of 220-230w.

The wind turbine system is a 15kw generator. After being accumulated in a battery bank, electrical energy is treated by an inverter/charger in order to produce an appropriate output current for computing and networking devices.

User applications are running on multiple Dell Poweredge R710 systems, hosted by a rack mount structure in an outdoor climate-controlled enclosure.

Within each node, servers are linked by a local network, which is connected then to the core network through GE transceivers.

Data flows are transferred among GSN nodes over dedicated circuits (like light paths or P2p links), tunnels over Internet or logical IP networks.

The Montreal GSN node plays a role of a manager (hub node) that opportunistically sets up required connectivity for Layer 1

then pushes Virtual machines (VMS) or software virtual routers from the hub to a sun or wind node (spoke node) when power is available.

The VMS are used to run user applications, particularly heavy-computing services. Based on this testbed network, experiments and research are performed targeting cloud management algorithms and optimization of the intermittently-available renewable energy sources.

The cloud management solution developed in order to run the GSN enables the control of a large number of devices of different layers.

the proposed solution aims at distributing user-oriented services Fig. 3. Layered GSN and Cloud computing Architectures Renewable Energy Provisioning for ICT Services in a Future Internet 425 regardless of the underlying

Such a management approach is essential for data center migration across a wide area-network network, because the migration must be achieved in a timely manner

which is a new software platform specific for dealing with the delivery of computing infrastructure 5. Figure 3 compares the layered architecture of the GSN with a general architecture of a cloud comprising four layers.

The GSN Data plane corresponds to the System level, including massive physical resources, such as storage servers and application servers linked by controlled circuits (i e.,

, lightpaths. The Platform Control plane corresponds to the Core Middleware layer, implementing the platform level services that provide running environment enabling cloud computing

and networking capabilities to GSN services. The Cloud Middleware plane corresponds to the User-level Middleware,

providing Platform as a service capabilities based on Iaas Framework components 5. The top Management plane or User level focuses on application services by making use of services provided by the lower layer

services. 4 Virtual Data center Migration In the GSN project, we are interested in moving a virtual data center from one node to another.

Such a migration is required for large-scale applications running on multiple servers with a high density connection local network.

The migration involves four steps: i) Setting up a new environment (i e.,, a new data center) for hosting the application with required configurations,

ii) Configuring network connection, iii) Moving VMS and their running state information through this high speed connection to the new location,

and iv) Turning off computing resources at the original node. Indeed, solutions for the migration of simple applications have been provided by many ICT operators in the market.

However, large scale data centers require arbitrarily setting their complex working environments when being moved. This results in a reconfiguration of a large number of servers and network devices in a multi-domain environment.

Fig. 4. Iaas Framework Architecture Overview 426 K. K. Nguyen et al. In our experiments with an online interactive application like Geochronos 7 each VM migration requires 32mbps bandwidth

in order to keep the service live during the migration, thus a 10 Gbit/s link between two data centers can transport more than 300 VMS in parallel.

Given that each VM occupies one processor and that each server has up to 16 processors,

20 servers can be moved in parallel. If each VM consumes 4gbyte memory space, the time required for such a migration is 1000sec.

The migration of data centers among GSN nodes is based on cloud management. The whole network is considered as a set of clouds of computing resources

which is managed using the Iaas Framework 5. The Iaas Framework include four main components: i) Iaas Engine used to create model and devices interactions abstractions,

ii) Iaas Resource used to build web services interfaces for manageable resources, iii) Iaas Service serves as a broker

which controls and assigns tasks to each VM, and iv) Iaas Tool provides various tools

and utilities that can be used by the three previous components (Figure 4). The Engine component is positioned at the lowest level of the architecture

Capabilities can contribute to a resources Business, Presentation or Data Access Tier. The Tool component provides additional services, such as persistence,

OSGI (Open Services Gateway initiative) is a Java framework for remotely deployed service applications, which provides high reliability, collaboration, large scale distribution and wide-range of device usage.

Through a Web interface, users may determine GHG emission boundaries based on information providing VM power and their energy sources,

and converges server and network virtualizations. Whilst most of cloud management solutions in the market focus particularly on computing resources,

Iaas Framework components can be used to build network virtualized tools 6 10, which provides for a flexible set of data flows among data centers.

The ability of incorporating third-party power control components is also an advantage of the Iaas Framework.

Renewable Energy Provisioning for ICT Services in a Future Internet 427 5 Federated Network GSN takes advantage of the virtualization to link virtual resources together to span multiple cloud and substrate types.

An orchestration middleware is built to federate clouds across domains, coordinate user registration, resource allocation, stitching,

and leverage and interoperate with software outside of the GSN. Along with the participation of international nodes, there is an increasing need of support for dynamic circuits on GSN

the client will contact firstly an energy-aware router in order to get an appropriate VM for his service.

The router will look for a VM which is optimal in terms of GHG emission, i e.,, the one which is powered by a green energy source.

in order to move a VM to a greener data center. The process is as follows (Figure 5: i) Copy VM memory between old and new locations,

iii) Router B receives the ARP and sends the message to the client, iv) New routing entry is installed in router B for the VM,

and v) New routing entry is added in router A. 428 K. K. Nguyen et al. In our design, the GSN is provided with a component called the Federation Stitcher

which is responsible for establishing connection among domains, and forwarding user requests to appropriate data centers. The big picture of the GSN network management solution is shown in Figure 6. The heart of the network is the programmable Federation Stitcher

which accepts connections from service users through Internet. This point is powered by green sustainable energy, i e.,

, hydroelectricity. It links user requests to appropriate services provided by data centers distributed across the network.

Each data center is represented by a virtual instance, including virtual servers and virtual routers and/or virtual switches interconnecting the servers.

Such a virtual data center can be hosted by any physical network node, according to the power availability.

There is a domain controller within each data center or a set of data centers sharing the same network architecture/policy.

User requests will be forwarded by the Federation Stitcher to the appropriate domain controller. When a VM or a data center is migrated

the new location will be registered with the Federation Stitcher then user requests are tunneled to the new domain controller.

Fig. 6. Overview of GSN network management solution Renewable Energy Provisioning for ICT Services in a Future Internet 429 6 Conclusion In this chapter, we have presented a prototype of a Future

Internet powered only by green energy sources. As a result of the cooperation between Europe and North america researchers, the Greenstar Network is a promising model to deal with GHG reporting

and carbon tax issues for large ICT organizations. Based on the Mantychore FP7 project a number of techniques have been developed

in order to provision renewable energy for ICT services worldwide. Virtualization techniques are shown to be the most appropriate solution to manage such a network

and to migrate data centers following green energy source availability, such as solar and wind. Our future work includes research on the quality of services hosted by the GSN and a scalable resource management.

Acknowledgments. The authors thank all partners for their contribution in the GSN and Mantychore FP7 projects.

Open Access. This article is distributed under the terms of the Creative Commons Attribution Noncommercial License

Converged Optical Network Infrastructures in Support of Future Internet and Grid Services Using Iaas to Reduce GHG Emissions.

Extending the Argia software with a dynamic optical multicast service to support high performance digital media.

HEANET website, http://www. heanet. ie/12. NORDUNET website, http://www. nordu. net 13. Moth, J.:

GN3 Study of Environmental Impact Inventory of Greenhouse Gas Emissions and Removals NORDUNET (9/2010) 14.

IBBT Website, http://www. ibbt. be/16. Reservoir FP7, http://www. reservoir-fp7. eu/J. Domingue et al.

) Future Internet Assembly, LNCS 6656, pp. 431 446,2011. The Author (s). This article is published with open access at Springerlink. com. Smart Cities and the Future Internet:

Towards Cooperation Frameworks for Open Innovation Hans Schaffers1, Nicos Komninos2, Marc Pallot3, Brigitte Trousse3, Michael Nilsson4, Alvaro Oliveira5 1 ESOCE Net hschaffers

and validating Future Internet-enabled services. Based on an analysis of the current landscape of smart city pilot programmes, Future Internet experimentally-driven research

and projects in the domain of Living Labs, common resources regarding research and innovation can be identified that can be shared in open innovation environments.

Smart Cities, Future Internet, Collaboration, Innovation Ecosystems, User Co-Creation, Living Labs, Resource Sharing 1 Introduction The concept of smart cities has attracted considerable

The Internet and broadband network technologies as enablers of e-services become more and more important for urban development

and user-driven innovation ecosystems to boost Future Internet research and experimentation for user-driven services and how they can accelerate the cycle of research,

This paper pays particular attention to collaboration frameworks which integrate elements such as Future Internet testbeds

while also encompassing peripheral and less developed cities. It also emphasises the process of economic recovery for welfare and well-being purposes.

Secondly, this characterisation implicitly builds upon the role of the Internet and Web 2. 0 as potential enablers of urban welfare creation through social participation, for addressing hot societal challenges, such as energy efficiency, environment

and user communities Table 1. Three perspectives shaping the landscape of Future Internet and City Development Future Internet Research Cities and Urban Development User-Driven Innovation

researchers as co-creators Priorities Future Internet technical challenges (e g. routing, scaling, mobility) Urban development Essential infrastructures Business creation User-driven open innovation Engagement of citizens Resources Experimental facilities Pilot environments Technologies Urban

collaborative innovation Smart Cities and the Future Internet 433 for experimentation on Future Internet technologies and e-service applications.

Common, shared research and innovation resources as well as cooperation models providing access to such resources will constitute the future backbone of urban innovation environments for exploiting the opportunities provided by Future Internet technologies.

in order to explore the conditions for rising to this challenge (see Table 1). The first perspective of Future Internet research

However, a wide gap exists between the technology orientation of Future Internet research and the needs and ambitions of cities.

A key challenge is the development of cooperation frameworks and synergy linkages between Future internet research, urban development policies and open userdriven innovation.

exploratory and participative playground combining Future Internet push and urban policy pull in demand-driven cycles of experimentation and innovation.

Living Labdriven innovation ecosystems may evolve to constitute the core of 4p (Public-Private-People-Partnership) ecosystems providing opportunities to citizens

explore, experiment and validate innovative scenarios based on technology platforms such as Future Internet experimental facilities involving SMES and large companies as well as stakeholders from different disciplines.

Section 2 addresses challenges for cities to exploit the opportunities of the Future Internet and of Living Lab-innovation ecosystems How methodologies of Future Internet experimentation and Living Labs could constitute the innovation ecosystems of smart cities is discussed in section 3. Initial examples of such ecosystems

and related collaboration models are presented in section 4. Finally, section 5 presents conclusions and an outlook. 2 City and Urban Development Challenges In the early 1990s the phrase"smart city"was coined to signify how urban development was turning towards technology,

innovation and globalisation 6. The World Foundation for Smart Communities advocated the use of information technology to 434 H. Schaffers et al. meet the challenges of cities within a global knowledge economy 7. However,

the more recent interest in smart cities can be attributed to the strong concern for sustainability,

and to the rise of new Internet technologies, such as mobile devices (e g. smart phones), the semantic web, cloud computing,

and the Internet of things (Iot) promoting real world user interfaces. The concept of smart cities seen from the perspective of technologies

and components has some specific properties within the wider cyber, digital, smart, intelligent cities literatures.

It focuses on the latest advancements in mobile and pervasive computing, wireless networks, middleware and agent technologies as they become embedded into the physical spaces of cities.

The emphasis on smart embedded devices represents a distinctive characteristic of smart cities compared to intelligent cities

and web-based applications of collective intelligence 8, 9. Box: A New Spatiality of Cities-Multiple Concepts Cyber cities, from cyberspace, cybernetics, governance and control spaces based on information feedback, city governance;

Digital cities, from digital representation of cities, virtual cities, digital metaphor of cities, cities of avatars, second life cities, simulation (sim) city.

Intelligent cities, from the new intelligence of cities, collective intelligence of citizens, distributed intelligence, crowdsourcing, online collaboration, broadband for innovation, social capital of cities, collaborative learning

Smart cities, from smart phones, mobile devices, sensors, embedded systems, smart environments, smart meters, and instrumentation sustaining the intelligence of cities.

with the help of instrumentation and interconnection of mobile devices, sensors and actuators allowing real-world urban data to be collected

Smart city Smart Cities and the Future Internet 435 solutions are expected to deal with these challenges

1) the development of broadband infrastructure combining cable, optical fibre, and wireless networks, offering high connectivity and bandwidth to citizens and organisations located in the city,(2) the enrichment of the physical space and infrastructures of cities with embedded systems, smart devices, sensors,

and actuators, offering realtime data management, alerts, and information processing, and (3) the creation of applications enabling data collection and processing, web-based collaboration,

and actualisation of the collective intelligence of citizens. The latest developments in cloud computing and the emerging Internet of things, open data, semantic web,

and future media technologies have much to offer. These technologies can assure economies of scale in infrastructure

standardisation of applications, and turn-key solutions for software as a service, which dramatically decrease the development costs

while accelerating the learning curve for operating smart cities. The second task consists of initiating large-scale participatory innovation processes for the creation of applications that will run

Future media research and technologies offer a series of solutions that might work in parallel with the Internet of things and embedded systems, providing new opportunities for content management 12,13.

Media Internet technologies are at the crossroads of digital multimedia content and Internet technologies, which encompasses media being delivered through Internet networking technologies,

and media being generated, consumed, shared and experienced on the web. Technologies, such as content and context fusion, immersive multi-sensory environments, location-based content dependent on user location and context, augmented reality applications, open and federated

platforms for content storage and distribution provide the ground for new e-services within the innovation ecosystems of cities (see Table 2). Table 2. Media Internet technologies

and components for Smart Cities Solutions and RTD challenges Short term (2014) Mid term (2018) Longer term (2022) Content management tools Media Internet technologies Scalable multimedia

compression and transmission Immersive multimedia Collaboration tools Crowd-based location content; augmented reality tools Content and context fusion technologies Intelligent content objects;

large scale ontologies and semantic content Cloud services and software components City-based clouds Open and federated content platforms Cloud-based fully connected city Smart systems based on Internet of things Smart power management Portable systems Smart systems enabling integrated solutions e g. health

and care Software agents and advanced sensor fusion; telepresence Demand for e-services in the domains outlined in Fig. 1 is increasing,

but not at a disruptive pace. There is a critical gap between software applications and the provision of e-services in terms of sustainability and financial viability.

Not all applications are turned into e-services. Those that succeed in bridging the gap rely on successful business models that turn technological capabilities into innovations,

secure a continuous flow of data and information, and offer useful services. It is here that the third task for city authorities comes into play,

Standardisation would dramatically reduce the development and maintenance costs of e-services due to cooperation, exchange Smart Cities and the Future Internet 437 and sharing of resources among localities.

Open source communities may also substantially contribute to the exchange of good practices and open solutions.

such as IBM, Cisco, Microsoft, are involved strongly in and are contributing to shaping the research agenda.

linking smart cities with user-driven innovation, future Internet technologies, and experimental facilities for exploring new applications and innovative services.

and open public data up to developers as well as user communities. As the major challenge facing European cities is to secure high living standards through the innovation economy

and the knowledge economy overall. 3 Future Internet Experimentation and Living Labs Interfaces In exploring the role of Future Internet experimentation facilities in benefiting urban development as we move towards smart cities,

as well as the potential role of the‘Living Labs'concept in enriching experimentally-driven research on the Future Internet.

and experiments combining heterogeneous technologies that represent key aspects of the Future Internet. The considerable obstacles of complexity and unfamiliarity that are faced

when trying to explore the effects of new applications that bring future users the increasing power of the Future Internet have not yet been overcome.

although some interesting initiatives in that respect have started such as the Smart Santander project (services and applications for Internet of things in the city),

logistics and environment Iot-based services. A comparison of the role of users in FIRE facilities projects compared to Living Labs is presented in Table 3. Importantly,

The European commission has voiced its support for stronger user orientation in the future Internet facilities projects; not only users in terms of academic and industry researchers who will use these facilities for their research projects, but also end-users.

Table 3. User Role in FIRE and Living Labs Future Internet Experiments Living Labs Innovation Approach Controlled experiments Observing large-scale deployment and usage

we will now take a further look at Living Labs. The Web 2. 0 era has pushed cities to consider the Internet,

including mobile networks, as a participative tool for engaging citizens and tourists. Many initiatives have been launched by cities

Apart from the diversity of research streams and related topics for designing alternatives of the Internet of tomorrow, it becomes increasingly challenging to design open infrastructures that efficiently support emerging events and citizens'changing needs.

and playable city ser Smart Cities and the Future Internet 439 vices based on real-time digital data representing digital traces of human activity and their context in the urban space.

telecommunication networks reflect connectivity and the location of their users; transportation networks digitally manage the mobility of people and vehicles as well as products in the city,

Today, it is becoming increasingly relevant to explore ways in which such data streams can become tools for people taking decisions within the city.

such as open innovation and open business models 16, Web 2. 0 17 as well as Living Labs 18, a concept originating from the work of William Mitchell at MIT

Altogether, Future Internet experimental facilities, Living Labs and Urban development programmes form an innovation ecosystem consisting of users and citizens,

Private and People Partnership) ecosystem that provides opportunities to users/citizens to co-create innovative scenarios based on technology platforms such as Future Internet technology environments involving large enterprises

It appears that Future Internet testbeds could be enabling the co-creation of innovative scenarios by users/citizens contributing with their own content

public data. 4 Emerging Smart City Innovation Ecosystems As Table 4 illustrates, several FP7-ICT projects are devoted to research and experimentation on the Future Internet and the Internet of things within cities,

such as Smart Santander and, within the Iot cluster, ELLIOT. The CIP ICT-PSP programme has initiated several pilot projects dedicated to smart cities and Living Labs, some with a clear Future Internet dimension (Apollon

Periphèria, and to a less extent too, Open Cities and EPIC. Among the earlier projects with interesting aspects on the interface of Living Labs and Future Internet is C@R (FP6.

440 H. Schaffers et al. The Smart Santander project proposes an experimental research facility based on sensor networks which will eventually include more than 20,000 sensors,

considered as Iot devices. The architecture supports a secure and open platform of heterogeneous technologies.

The project is intended to use user-driven innovation methods for designing and implementing‘use cases'.

a map of sensor data available on smart phone) as well as urban waste management are two of the use cases from the Smart Santander project.

Internet services and sensor network in the city. www. smartsantander. eu ELLIOT (FP7-ICT, 2010.

Experimental Living Lab for Internet of things. Three Living Labs are involved. http://www. elliot-project. eu/Periphèria (CIP ICT-PSP, 2010.

Internet of things in Smart City. www. peripheria. eu Open Cities (CIP ICT-PSP, 2010. Public sector services.

The ELLIOT project (Experiential Living Lab for the Internet of things) represents a clear example of Living Labs

and Future Internet interaction, elaborating three Iot use cases in three different Living Labs. The first use case is dedicated to co-creation by users of green services in the areas of air

such as the regional institution for air measurement quality (Atmo PACA), the local research institute providing the Iot-based green service portal

the Internet Foundation for the New Generation (FING) facilitating user workshops, and a local SME providing data access from electric cars equipped with air quality sensors (VULOG) and a citizen IT platform (a regional Internet space for citizens in the NCA area).

The objectives of the Iot-based green services use case are twofold: to investigate experiential learning of the Iot in an open and environmental data context,

and to facilitate the co-creation of Smart Cities and the Future Internet 441 green services based on environmental data obtained via sensors.

Various environmental sensors will be used, such as fixed sensors from Atmo PACA in the NCA area, fixed Arduino-assembled sensors by citizens, mobile sensors, such as citizen-wired green watches or sensors installed on electric

vehicles. The backbone of the green services use case is based an Iot service portal which addresses three main Iotrelated portal services by allowing the user:

1) to participate in the collection of environmental data; 2) to participate in the co-creation of services based on environmental data;

and 3) to access services based on environmental data, such as accessing and/or visualising environmental data in real time.

Three complementary approaches have already been identified as relevant for the green services use case: participatory/usercentred design methods;

diary studies for Iot experience analysis, and coupling quantitative and qualitative approaches for portal usage analysis. In this context of an open innovation and Living Lab innovation ecosystem,

focus groups involving stakeholders and/or citizen may be run either online or face-to-face. The Periphèria project is among the Smart Cities portfolio of seven projects recently launched in the European commission ICT Policy Support Programme.

Their aim is to develop smart cities infrastructures and services in real-life urban environments in Europe.

and the Internet of things European Research Cluster (IERC) and can therefore be taken as a model of Smart Cities and Future Internet integration.

At the core of Periphèria lies the role of Living Labs in constituting a bridge between Future Internet technology push

and Smart City application pull, refocusing the attention on People In places to situate the human-centric approach within physical urban settings.

which the integration of Future Internet infrastructures and services occurs as part of a discovery-driven process.

Participation is at the heart of this bottom-up approach to Future Internet technology integration, whereby Future Internet research adopts a competitive offer stance to prove its added value to users.

Platform and service convergence is promoted by the use of serious games that engage citizens and users in the process of discovering the potential of Future Internet technologies

and the possible sustainable scenarios that can be built upon them. Serious gaming thus constitutes a mechanism to enhance participation

in addition, they constitute a monitoring and governance platform for increasing self-awareness of the changes brought about by the adoption of Future Internet technologies.

This approach draws on and integrates Future Internet technologies (such as augmented reality services for the appreciation of cultural heritage) with networks of video-cameras used to monitor public spaces.

and prioritisation of the cultural heritage in their city and also to an exploration of the privacy and security issues that are central to the acceptance and success of Future Internet services for the safety of urban environments.

and workstyles made possible by Future Internet technologies. In addition, it shows how the Future Internet is a mixture of technologies and paradigms with overlapping implementation time-frames.

While the deployment of IPV6 networks may be a medium-term effort other Future Internet paradigms such as cloud services and camera and sensor networks can be considered as already operational.

The discovery-driven arena settings in Periphèria are guiding the development of Living Lab-convergent service platforms that bring these technologies together into integrated,

dynamic co-creation environments that make up a Smart City. These projects examples provide initial examples of collaboration models in smart city innovation ecosystems, governing the sharing and common use of resources such as testing facilities, user groups

Still, many Smart Cities and the Future Internet 443 issues need to be clarified such as how the different research and innovation resources in a network,

such as specific testing facilities, tools, data and user groups, can be made accessible and adaptable to specific demands of any research and innovation projects.

The ELLIOT project is an example of a Future Internet research and innovation project embedded in regional and even national innovation policy.

and developers. 444 H. Schaffers et al. 5 Conclusions and Outlook In this paper we explored the concept of smart cities as environments of open

and validating Future Internet-enabled services. Smart cities are enabled by advanced ICT infrastructure contributed to by current Future Internet research and experimentation.

Such infrastructure is one of the key determinants of the welfare of cities. Other determinants of the welfare of cities will be important as well:

Based on an analysis of challenges of smart cities on the one hand and current projects in the domain of Future Internet research and Living Labs on the other,

common resources for research and innovation can be identified, such as testbeds, Living Lab facilities, user communities, technologies and know-how, data,

One layer focuses on the actual resources within the Future Internet research and innovation process

e g. the use of Living Lab facilities and methods in experimenting on Future Internet technologies,

Initial examples of resource sharing appear in making user communities available for joint use with Future Internet facilities (e g. the TEFIS project),

and in making accessible Future Internet facilities for developing and validating Iot-based service concepts and applications through Living Labs approaches for smart cities (e g. the Smartsantander and ELLIOT projects).

The Future Internet constitutes both a key technology domain and a complex societal phenomenon. Effective, user driven processes of innovation, shaping and application of Smart Cities and the Future Internet 445 Future Internet technologies in business and society are crucial for achieving socioeconomic benefits.

A key requirement emphasised in this paper is how, within an environment of open innovation in smart cities and governed by cooperation frameworks,

Smart Cities, Fast Systems, Global networks. Rowman & Littlefield, New york (1992) 7. WFSC: Smart Communities, http://www. smartcommunities. org/about. htm 8. Komninos, N.:

IBM Journal of Research & development 53 (3), 338 353 (2009) 11. European commission: Growing Regions, Growing Europe:

Future Media Internet: Research challenges and road ahead. DG Information Society and Media, Luxembourg, Publications Office of the European union (2010) 13.

Future Internet Research and Experimentation (September 2010) 16. Chesbrough, H. W.:Open Innovation: The New Imperative for Creating

Web Squared: Web 2. 0 Five Years On. Special report, Web 2. 0 Summit, Co-produced by O'reilly & Techweb (2009) 18.

European commission, DG INFSO: Advancing and Applying Living Lab Methodologies (2010) 19. Ballon, P.,Pierson, J.,Delaere, S.,et al.:

Test and Experimentation Platforms for Broadband Innovation. IBBT/VUB-SMIT Report (2005) 446 H. Schaffers et al. 20.

) Future Internet Assembly, LNCS 6656, pp. 447 462,2011. The Author (s). This article is published with open access at Springerlink. com. Smart Cities at the Forefront of the Future Internet José M. Hernández-Muñoz1, Jesús Bernat Vercher1, Luis

Muñoz2, José A. Galache2, Mirko Presser3, Luis A. Hernández Gómez4, and Jan Pettersson5 1 Telefonica I+D, Madrid, Spain {jmhm, bernat}@ tid. es 2 University of Cantabria, Santander, Spain {luis, jgalache}@ tlmat. unican

. es 3 Alexandra Institute, Aahrus, Denmark mirko. presser@alexandra. dk 4 Universidad Politécnica Madrid, Spain luisalfonso. hernandez@upm. es 5

Smart cities have been pointed recently out by M2m experts as an emerging market with enormous potential,

which is expected to drive the digital economy forward in the coming years. However, most of the current city and urban developments are based on vertical ICT solutions leading to an unsustainable sea of systems and market islands.

In this work we discuss how the recent vision of the Future Internet (FI), and its particular components, Internet of things (Iot) and Internet of Services (Ios), can become building blocks to progress towards a unified urban-scale ICT platform transforming a Smart City into an open innovation platform.

Moreover we present some results of generic implementations based on the ITU-T's Ubiquitous Sensor Network (USN) model.

the infrastructure level (Iot to support the complexity of heterogeneous sensors deployed in urban spaces),

and at the service level (Ios as a suit of open and standardized enablers to facilitate the composition of interoperable smart city services).

Smart Cities, Sensor and Actuator Networks, Internet of things, Internet of Services, Ubiquitous Sensor Networks, Open, Federated and Trusted innovation platforms, Future Internet. 1 Introduction At a holistic level,

control, and monitor complex interdependent systems of dense urban life 3. Therefore in the design of urban-scale ICT platforms,

three main core functionalities can be identified: Urban Communications Abstraction. One of the most urgent demands for sustainable urban ICT developments is to solve the inefficient use (i e. duplications) of existing or new communication infrastructures.

and will enable data transfer services agnostic to the underlying connection protocol. Furthermore, a major challenge in future urban spaces will be how to manage the increasing number of heterogeneous and geographically dispersed machines

so that data and information could be shared among different applications and services at global urban levels.

and citizens) will be able to conceive new innovative solutions to interact Smart Cities at the Forefront of the Future Internet 449 with

In this work we advocate that this technological leap can be done by considering Smart Cities at the forefront of the recent vision of the Future Internet (FI.

Although there is no universally accepted definition of the Future Internet, it can be approached as a socio-technical system comprising Internet-accessible information and services, coupled to the physical environment and human behavior,

and supporting smart applications of societal importance 4. Thus the FI can transform a Smart City into an open innovation platform supporting vertical domain of business applications built upon horizontal enabling technologies.

The Internet of things (Iot: defined as a global network infrastructure based on standard and interoperable communication protocols where physical and virtual things are integrated seamlessly into the information network 5. The Internet of Services (Ios):

flexible, open and standardized enablers that facilitate the harmonization of various applications into interoperable services as well as the use of semantics for the understanding,

combination and processing of data and information from different service provides, sources and formats. The Internet of People (Iop:

envisaged as people becoming part of ubiquitous intelligent networks having the potential to seamlessly connect, interact and exchange information about themselves and their social context and environment.

At this point, it is important to highlight a bidirectional relationship between the FI and Smart Cities:

Section 2 discusses how major components of the Future Internet, namely Iot and Ios, can be essential building blocks in future Smart Cities open innovation platforms.

Several technical details related to the development of next generation urban Iot platforms are outlined in Section 3. Section 4 discusses the need for realistic urban-scale open

and federated experimental facilities, and presents most relevant current initiatives, with special attention to the Smartsantander EU Project.

Finally, conclusions and future challenges are given in Section 5. 2 Iot and Ios as ICT Building blocks for Smart Cities In the analysis from Forrester research 9 on the role that ICT will play in creating the foundation for Smart Cities,

a smart city is described as one that uses information and communications technologies to make the critical infrastructure components and services of a city administration, education, healthcare, public safety, real estate, transportation and utilities more aware, interactive and efficient.

Advanced location based services, social networking and collaborative crowdsourcing collecting citizens'generated data. By analyzing these different Smart Cities application scenarios, together with the need of a broadband communication infrastructure that is becoming,

Smart Cities at the Forefront of the Future Internet 451 Recent advances in Sensors and Actuator Networks (SAN) are stimulating massive sensor networks deployments, particularly for the previously described urban application areas.

Therefore Iot, essential to the FI, can be invaluable to provide the necessary technological support to manage in a homogeneous and sustainable way the huge amount of sensor

Ios evolution must be correlated undoubtedly with Iot advances. Otherwise, a number of future Smart City services will never have an opportunity to be conceived due to the lack of the required links to the real world.

and challenges of implementing Iot and Ios at the city scale. Starting with the benefits of Iot technologies, they are twofold:

on the one hand they can increase the efficiency, accuracy and effectiveness in operation and management of the city's complex ecosystem and, on the other,

they can provide the necessary support for new innovative applications and services (the city as an Open Innovation Platform).

in order to develop cross-domain Next Generation (NG) Iot platforms suitable to different usage areas and open business models to improve market dynamics by involving third parties in the value chain (SMES).

Some of the essential functionalities identified as required for NG Iot platforms comprise the support for horizontality, verticality, heterogeneity, mobility, scalability,

Cross-domain NG Iot platforms may foster the creation of new services taking advantage of the increasing levels of efficiency attained by the reuse of deployed infrastructures.

Considering now the Ios, it must be stressed that it is recognized widely (see for example 12) that the real impact of future Iot developments is tied heavily to the parallel evolution of the Ios. So,

a Smart City could only become a true open innovation platform through the proper harmonization of Ios and Iot.

There can be a long list of potential benefits for Smart Cities'services relaying on the same basic sensed information and a suite of application enablers (i e. from sensor data processing applications,

to enablers for accessing multimedia mobile communications or social networks, etc.).Thus the integration of innovative principles and philosophy of Ios will engage collective end-user intelligence from Web 2. 0

and Telco 2. 0 models that will drive the next wave of value creation at urban scales,

The technological challenge of developing the Ios has been assumed at EU level, and actions are being initiated to overcome the undesirable dissociation between technological

and deployed Iot platforms). In that way, an increasing number of Smart Cities'services could be searched,

and composed (following Web 2. 0/Telco2. 0 principles and including Qos, trust, security, and privacy) in a standard, easy and flexible way.

Furthermore this will enable future urban models of convergent IT/Telecom/Content services, Machine to machine-Machine (M2m) services,

Experimental Testbeds Ad hoc WSN Deployments Iotresources (sensor & actuator networks) Ios resources Testbed 1 USN-Enabler Service 1 Adaptation& Homogeneization Testbed

Control Layer GSDP SDP Entity exposure Service exposure Ios federation level Iot federation level NGN/Telco2. 0 Web2. 0 Service

2 Iot Service Service ntestbed n Domain n A&h Control Layer A&h Control Layer Other Enablers Domain 1 A&h Control

Layer Fig. 1. Global Service Delivery Platform (GSDP) integrating Iot/Ios building blocks 3 Developing Urban Iot Platforms At present, some works have been reported of practical implementations

in order to develop Iot platforms inspired by the Ubiquitous Sensor Networks concept from the ITU-T USN Standardization Group 21.

where data is binding the different dimensions, as most aspects are related closely (e g. environment and traffic, both of them to health, etc.).

Smart Cities at the Forefront of the Future Internet 453 3. 1 USN Functionalities The main goal of a USN platform is to provide an infrastructure that allows the integration of heterogeneous

many Smart City services will rely on continuously generated sensor data (for example for energy monitoring, video surveillance or traffic control.

This functionality will provide a repository where observations/sensors'data are stored to allow later retrieval or processing,

to extract information from data by applying semantic annotation and data linkage techniques. Publish-Subscribe-Notify:

Unified communication protocol: given the extension of an urban area, several standards can coexist to communicate sensors and sensor networks (Zigbee, 6lowpan, ISA-100.11. a, xdsl, GPRS, etc.).

Services should be agnostic to the communication protocol used. The platform should provide access to the information regardless the particular underlying communication protocol used. 454 J. M. Hernández-Muñoz et al.

Horizontally layered approach: The platform should also be built following a layered approach, so services and networks are decoupled

This capability will allow a seamless link between Iot and Ios, as discussed in Section 2. Also relevant will be the definition of open APIS,

so that USN platforms could provide support for third-party's agents interested in the deployment of different Smart City services,

and different business processes. 3. 2 USN Architecture for Urban Iot Platforms While the new wave of Next Generation Iot platforms are expected to be defined by initiatives and projects like Iot-A 23,

the IERC cluster 24 or the emerging PPP Iot Core Platform Working group discussion 25, multiple different approaches for First Generation Iot-platforms are currently being implemented.

In essence, many of them are realizations of the described ITU-T's model. For reference on the current state of the technology, this Section describes a practical USN platform implementation (more details can be found in 22),

integrated into The next Generation Networks Infrastructures 35, as one of the most remarkable currently reported solutions for advanced Iot platforms.

As shown in Figure 2, a functional specialization of the building blocks has been applied in this work. USN-Management USN-Enabler Sensor Networks IMS User Equipment USN-Gateway SIP Services Web Services Configuration AAA Devicemanagement Application/Service

Layer Control Layer Access Layer Service Protocol Adapter Notification Entity (NE) Sensor Tasking Entity (STE) Catalog Entity (CE) Sensor Description Entity

(SDE) Observation Storage Entity (OSE) Messages& Data format Adapter Communication protocol Adapter Fig. 2. High-level Architecture of a USN Iot Platform Smart Cities at the Forefront

of the Future Internet 455 As sketched in the figure, the USN platform is based on two components,

This approach is inspired by the Open Geospatial Consortium (OGC) Sensor Web Enablement (SWE) activity 26.

Its goal is the creation of the foundational components to enable the Sensor Web concept,

where services will be capable to access any type of sensors through the web. This has been reflected by a set of standards used in the platform (Sensorml, Observation & Measurements, Sensor Observation Service, Sensor Planning Service, Sensor Alert Service and Web Notification Service 26.

Besides the SWE influence, the USN-Enabler relays on existing specifications from the OMA Service Environment (OSE) 27 enablers (such as presence, call conferencing, transcoding, billing, etc..

The USN-Gateway represents a logical entity acting as data producers to the USNENABLER that implements two main adaptation procedures to integrate physical or logical Sensor and Actuator Networks (SANS:

Communication protocol Adaptation. As a connection point between two networks (sensors networks deployed throughout the city and the core IP communication network),

the main responsibility is to provide independence from the communication protocol used by the sensor networks.

Sensor Data format Adaptation. This functionality is intended to provide USNENABLER both Sensorml (meta-information) and O&m (observation & measurements) data from specific SANS data (i e.

Zigbee. Adaptation and Homogenization are two key requirements for the USN Platform aiming at its integration with different Smart Cities'testbeds and experimental deployments.

They are also essential requirements for a successful seamless integration, and the proper basement for the new heterogeneous sensor network infrastructures needed to enable an evolving FI based on the Iot and Ios paradigms.

Functionalities required to support services are offered both in synchronous and asynchronous mode by the USN-Enabler through the following entities:

The Notification Entity (NE) is the interface with any sensor data consumer that require filtering

or information processing over urban-generated data. The main functionalities provided by this entity are the subscription (receive the filter that will be applied

like for example a request to gather data, without the need to wait for an answer.

when the desired data gets available it will receive the corresponding alert. This is mainly used for configuration and for calling actuators.

The Service Protocol Adapter (SPA) provides protocol adaptation between the Web Services and SIP requests and responses.

simulation results can only give very limited information about the feasibility of an algorithm or a protocol in the field.

which the necessary infrastructure of a Smart City will rely on technologies of the Iot. The resulting scale and heterogeneity of the environment makes it an ideal environment for enabling the above mentioned broad range of experi Smart Cities at the Forefront of the Future Internet 457 mentation needs.

Furthermore, a city can serve as an excellent catalyst for Iot research, as it forms a very dense techno-social ecosystem.

Cities can act as invaluable source of challenging functional and nonfunctional requirements from a variety of problem and application domains (such as vertical solutions for the environment control and safety

and end-users that are required for testing of Iot as well as other Future Internet technologies for market adoption.

Node WISELIB User Developed App Tinyos Contiki Sunspot Opencom Middleware Mobility support Horizontal support Federation support Security, Privacy and Trust Fig

and aims at creating a unique-in-the-world European experimental test facility for the research and experimentation of architectures, key enabling technologies, services and applications for the Iot.

and Internet researchers to validate their cutting-edge technologies (protocols, algorithms, radio interfaces, etc.).Several use cases are currently under detailed analysis for their experimental deployment taking into account relevant criteria from local and regional authorities.

Tourism information in different parts of the city through mobile devices using visual and interactive experiences and in different languages.

Smart Cities at the Forefront of the Future Internet 459 Video monitoring for traffic areas, beach areas and specific events in public places, such as airports, hotels, train stations, concerts and sport stadiums.

research and service oriented initiatives on both Iot and Ios areas as WISEBED 25, SENSEI 8 and the USN Iot Platform (presented in Section 3) including Web 2. 0 and Telco 2. 0 design principles.

Additionally, the requirements elicitation process in Smartsantander will also consider the following viewpoints: the FIRE testbed user, the service provider, the service consumers (citizens), the Smartsantander facility administrators,

the Smartsantander middleware) that provide the functionality described by these requirements and is expected to accommodate additional requirements coming up from the different smart city services (use cases).

i) Access control and IOT Node Security subsystem, ii) Experiment Support Subsystem, iii) the Facility Management Support Subsystem,

The architectural reference model also specifies, for each sub-system, required component deployments on the Iot nodes,

This will not only reduce the technical and societal barriers that prevent the Iot concept to become an everyday reality

at least and not less important, provide the means to guarantee its day by day maintenance. 5 Conclusions Future Internet potential,

through Iot and Ios, for creating new real-life applications and services is huge in the smart city context.

First time success of large Iot deployments is jeopardized seriously by the lack of testbeds of the required scale,

validation of their viability as candidate solutions for real life Iot scenarios. At present, some practical implementations of advanced USN platforms 22 have been demonstrated successfully in real deployments for smart metering services, smart places scenarios,

Ongoing activities are extending its scope to broader M2m scenarios, and large scale deployments for experimental smart urban spaces.

providing the key components required to intertwining Iot and Ios worlds. Referred Iot USN platform is currently being evolved with the addition of new capabilities

and integrated within other components being developed previously by the EU projects SENSEI 8 and WISEBED 33 to implement a city scale infrastructure for Iot technologies experimentation within the Smartsantander project.

In this project, a large infrastructure of about 20,000 Iot devices is addressed. Currently, the deployment of the first 2, 000 sensors in the urban environment is been carried.

Nontechnical aspects are also of a big importance. The cardinality of the different stakeholders involved in the smart city business is so big that many nontechnical constraints must be considered (users, public administrations

vendors, etc..In this sense, what may be evident from a purely technique perspective it is not so clear

Nowadays, there are no field experiences across the world allowing assessing, in the short term, the behavior of massive wireless sensor deployments.

Future Internet applications relevant for smart cities, an ICT application area example: smart & proactive energy management, Open Innovation by FI-enabled services, Brussels, 15 january (2010) Smart Cities at the Forefront of the Future Internet 461 4. Position Paper:

Research Challenges for the Core Platform for the Future Internet. In: M. Boniface, M. Surridge, C. U (Eds.

http://ec. europa. eu/information society/activities/foi/library/docs/fippp-research-challenges-for-coreplatform-issue-1-1. pdf 5. Sundmaeker, H

Vision and Challenges for Realising the Internet of things, CERP-Iot, March 2010. European commission, Brussels (2010) 6. Future Internet Assembly 2009, Stockholm, Sweden (November 2009), http://ec. europa. eu/information society/activities/foi/library

/docs/fi-stockholm-report-v2. pdf 7. The European Network of Living Labs, http://www. openlivinglabs. eu/8. SENSEI Integrating the Physical with the Digital

State of the art Sensor Frameworks and Future Internet (D3. 1). Technical report (2008) 9. Belissent, J.:Getting Clever About Smart Cities:

Towards a Future Internet Public Private Partnership, Usage Areas Workshop, Brussels, 3 march (2010), http://ec. europa. eu/information society/activities/foi/events/fippp3

Real world Internet (RWI) Session, FIA meeting, Prague (May 2009), http://rwi. future-internet. eu/index. php/RWISESSION PRAGUE 13. COM:

A public-private partnership on the Future Internet. Brussels, 28 october (2009), http://ec. europa. eu/information society/activities/foi/library/docs/fi-communication en. pdf 14.

DG INFSO Task force on the Future Internet Content. Draft Report of the Task force on Interdisciplinaryresearch Activities applicable to the Future Internet, Version 4. 1 of 13.07.2009 (2009), http://forum. future-internet. eu 15.

NESSI Strategic Research Agenda, http://www. nessi-europe. com/files/Researchpapers/NESSI SRA VOL 3. pdf 16. Gluhak, A.,Bauer, M.,Montagut, F.,Stirbu, V.,Johansson, M.,Bernat-Vercher, J.,Presser, M.:

Towards an architecture for a Real world Internet. In: Tselentis, G.,et al. eds.)) Towards the Future Internet, IOS Press, Amsterdam (2009) 17.

Fisher, S.:Towards an Open Federation Alliance. The WISEBED Consortium. Lulea, July 2nd, 2009.22. In: Balazinska, M.,et al.

) Data Management in the Worldwide Sensor Web. IEEE PERVASIVE computing, April-June (2007) 18. Panlab Project, Pan European Laboratory Infrastructure Implementation, http://www. panlab. net/fire. html 19.

Global service delivery platform (GSDP) for the future internet: What is it and how to use it for innovation?,

http://services. future-internet. eu/images/d/d4/Report GSDPPANEL-FISO-FIA-Madrid-draft%2breqs. pdf 20.

Future Internet Assembly, Meeting Report, Madrid, Spain, 9th 10th december (2008), http://ec. europa. eu/information society/activities/foi/library/docs/madrid

-conference-report-v1-1. pdf 21. ITU TSTAG: A preliminary study on the Ubiquitous Sensor Networks.

Internet of things Architecture project, http://www. iot-a. eu/public/front-page 462 J. M. Hernández-Muñoz et al. 24.

Iot European Research Cluster, http://www. internet-of-things-research. eu/25. White paper on the FI PPP definition (Jan. 2010), http://www. future-internet. eu/fileadmin/initiative documents/Publications/White paper/EFII White paper 2010 public. pdf 26.

Botts, M.,Percivall, G.,Reed, C.,Davidson, J.:OGC Sensor Web Enablement: Overview and High Level Architecture, Open Geospatial Consortium Inc. White paper Version 3 (2007) 27.

OMA Service Environment Archive, http://www. openmobilealliance. org/technical/release program/ose archive. aspx 28. Oulu Smart City, http://www. ubiprogram. fi/29.

WISEBED-Wireless Sensor Network Testbeds, http://www. wisebed. eu 34. Onelab2, ONELAB project, http://www. onelab. eu/35.

< Back - Next >

Overtext Web Module V3.0 Alpha
Copyright Semantic-Knowledge, 1994-2011