Synopsis: Ict: Computer:

Moreover, the use of computers has expanded to almost the entirety of these companies (99.2%).%)In turn, 87.3%had a Local area network (LAN) installed, and

-Computers 99.2 99.1 99.5 99.8 -Local area network 87.3 85.9 95.0 97.7 -Wireless Local area network 61.9 59.5 74.2 81.1

the internet for company use. 36.4%of these were laptop computers, and 49%were smartphones or PDA phones

server (54.7 %53.4%of the companies that used Cloud computing did so by paying any service existing in

servers of shared services suppliers The main reasons that limited the companies using this service were the uncertainty about

72.3%of companies with fewer than 10 employees had computers, and 24.4%had a Local

-Computers 71.6 72.3 -Local area network 24.0 24.4 -Wireless Local area network 16.4 17.6 -Internet connection 65.7 67.7 -Broadband (fixed or mobile) Internet connection (1) 98.5 99.0

Communication Technologies based on Open sources (CENATIC Moreover, there is another partnership agreement with the Ministry of Industry, Tourism and

large companies and was implemented by means of computer-assisted telephone interviewing. Data collection was done over a 2 month period during September-October

companies NACE code principal-6201 (Activities to develop custom software (software -oriented client), 20.9%CAEN 6202 (consultancy activities information technology), 1. 2

core of social innovations is this intention to create something better. But while these motivations might

the adoption of the core †content†of the innovation. After this initial period, their role is

their role is to allow the core innovation to fit into a new context. But sometimes adaptations

their role is to allow the core innovation to fit into a new context. But sometimes adaptations

and mobile devices, new ways of carrying out social innovation have become possible. Often this means the barriers to social innovation (e g. connectivity

knowledge and practices among the core competencies that are crucial in creating A Culture of

Microsoft Research, Cambridge, MA, USA Demetri Terzopoulos University of California, Los angeles, CA, USA Doug Tygar

SL 5 †Computer Communication Networks and Telecommuni -cations  The Editor (s)( if applicable) and the Author (s) 2011.

The core of this program will be a platform that implements and inte -grates new generic but fundamental capabilities of the Future Internet, such as interac

-tions with the real world through sensor/actuator networks, network virtualization and cloud computing, enhanced privacy and security features and advanced multimedia

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

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

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

multimedia communications, taking advantage of advanced software services, buying and selling, keeping in touch with family and friends, to name a few.

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

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

-tectural co-existence of new and legacy services and networks, via virtualisation of connectivity and computation resources and self management capabilities, by fully

The Internet has evolved from a remote access to mainframe computers and slow communication channel among scientists to the most important medium for informa

taking advantage of advanced software services, buying and selling, keeping in touch with family and friends, to name a few.

of nodes (computers, terminals mobile devices, sensors, etc. of the Internet will soon grow to more than 100 billion 1. Reliability, availability, and interoperability re

-ronments, network gaming, virtual worlds, etc. compared to the quantity and type of data currently exchanged over the Internet.

computers (e g.,, terminals, servers, etc. CPUS, etc. and handlers (software pro -grams/routines) that generate

and treat as well as query and access data •Storage of â€oedataâ€: refers to memory, buffers, caches, disks, etc.

and associated logical data structures •Transmission of â€oedataâ€: refers to physical and logical transferring/exchange of data

-nomenon will still cause thousands of cache servers to request the same docu -ments from the original site of publication

Software & Service Architectures & Infrastruc -tures, D4: Networked Enterprise & Radio frequency identification (RFID) and F5

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

ACM Computer Communications 33 (17), 2105†2115 (2010 19 Freedman, M.:Experiences with Coralcdn: A Five-Year Operational View.

ACM SIGCOMM Computer Com -munication Review 39 (5)( 2009 26 Eggert, L.:Quality-of-Service:

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

of new and legacy services and networks, via virtualisation of connectivity and computation resources and self management capabilities, by fully integrating

In-Network Clouds, Virtualisation of Resources, Self management Service plane, Orchestration plane and Knowledge plane 1 Introduction

-vices 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

-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 Inter

-mon physical substrate of communication nodes and servers managed by multiple infrastructure providers. By decoupling service providers from infrastructure provid

and execute on network hosts, devices, attachments, and servers Towards In-Network Clouds in Future Internet 21

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

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

2. 3 Virtualisation Plane Overview Virtualisation hides the physical characteristics 14,16 of the computing and net

-working resources being used, from its applications and users. 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 rout -ing processes by assigning different internal resources to each routing process;

alter -natively, multiple physical resources can appear as a single physical resource (e g when multiple switches are â€oestacked†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

This paper extends contemporary virtualisation approaches and aims at build -ing 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 re -sources into appropriate sets of virtual resources that can be organised by the Orches

the vspi and the vcpi (Virtualisation System Programming interface and Virtualisa -tion Component Programming interface, respectively. A set of control loops is

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

(and implicitly the AMS and DOC that are part of a given Orchestration Plane) to govern virtual resources,

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

physical resource by the vcpi of the Virtualisation Plane. The AMS sends abstract i e.,, device-independent) commands via the vcpi,

clients,(ii) the Context Processing (CP) Module which implements the core internal operations related to the context processing,(iii) the Context Information Base (CIB

in support of the Context Executive and Context Processor modules. Context is dis -tributed and replicated within the domain

they monitor hardware and software for their state, present their capabili -ties, or collect configuration parameters.

A monitoring mechanism and framework was developed to gather measurements from relevant physical and virtual resources

, 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

-erations such as getting the state of a server with its CPU or memory usage, getting

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

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.

, physical nodes and servers) subject to constraints determined by the Orchestration Plane. The Management Plane is de

It monitors the network and operational context as well as internal operational network state in order to assess if the network cur

-cal nodes and servers) subject to constraints determined by the Orchestration Plane Fig. 5. Autonomic Control Loops

-oped and issued as open source 10, which aims to create a highly open and flexible

Full design and implementation of all software platforms are presented in 10 •vcpi (Virtual Component Programming interface is the VP€ s main component deal

add powerful and flexible monitoring facilities to system clouds (virtualisation of networks and services. Lattice has a minimal runtime footprint

-tual entities, such as servers and routers •RNM (Reasoning and Negotiation Module), a core element of the KP, which me

-diates and negotiates between separate federated domains These In-Network Cloud platforms were integrated and validated on 2 testbeds ena

Testbed located in London consisting of 80 cores with a dedicated 10 Gbits/s infra

cores and linked by a dedicated 10 Gbits/s infrastructure. Validation and performance analysis are described fully in 13.

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

and service resources provided by an virtualisation environment. We have described also the management architectural and system model

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: Man

-aging the heterogeneity through introduction of homogeneous virtual resources and enabling programmability of the network elements.

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

Platforms and Software systems for an Autonomic Internet. IEEE Globecom 2010; 6-10 dec.,, Miami, USA (2010

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.:

-tures, and Protocols For Computer Communications (Karlsruhe, Germany, SIGCOMM †03 Karlsruhe, Germany, August 25†29,2003, pp. 3†10.

The International Journal of Computer and Tele -communications Networking 54 (5)( 2010 15. Galis, A.,Denazis, S.,Bassi, A.,Berl, A.,Fischer, A.,de Meer, H.,Strassner, J.,Davy, S.,Ma

IOS Press, Amsterdam (2009), http://www. iospress. nl /16. Berl, A.,Fischer, A.,De Meer, H.:

Using System Virtualization to Create Virtualized Net -works. Electronic communications of the EASST 17, 1†12 (2009

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 cur -rent trends in telecommunications, vendors prognosticate that mobile networks will suffer an immense traffic explosion in the packet switched domain up to year 2020

application areas (virtual reality experience sharing and gaming) will further boost this process and set new challenges to mobile networks.

deeply impregnated with social networking and virtual reality Even though video seems to be a major force behind the current traffic growth of

-tions from the core to the edge of the network results in sustainable network infra

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 sum -marized in the followings. Fig. 1 illustrates the evolution steps of the packet-switched

-nology in the core and backhaul network: Release 4 (2003) specified the Media gate -way 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

, 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, i e. the base

functions, 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

the core network entities. SIPTO enables per APN and/or per IP flow class based

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 central -ized, keeping user IP traffic anchored.

-tion of certain control functions still provided by the core. UFA represents the ulti -mate step toward flattening IP-based core networks, e g.,

, the EPC in 3gpp. The ob -jective 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

-dures of UFA, to reduce the number of HIP Base Exchanges in the access and core

-main in the core network. A good example is the Global HA to HA protocol 34, which

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 administration of the system of distributed MAS is done via a distributed Home Agent overlay table structure based on a Distributed Hash Table

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

-spondent node, a special information server is required in the network, which can also Flat Architectures:

by using Hi3 50 for core-level distribution of HIP signaling plane) are also feasible

directly connected to the IP core infrastructure. Therefore, they provide convenient and implicit interoperability between heterogeneous wireless technologies, and facili

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 distrib

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

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

Telecommunications Software and Systems Group †TSSG, Co. Waterford, Ireland {jmserrano, sdavy, mjohnsson, wdonnelly}@ tssg. org

-placing a plethora of proprietary hardware and software platforms with generic solu -tions supporting standardised development and deployment stacks

core Internet Protocols themselves We argue that service agnostic network design are no longer a way to achieve in

and outlook of this research. Finally some bibliography references supporting this research are included 2 Challenges for Future Internet Architectures

-ships between Network Virtualisation and Federation 16 21 22 23 and the rela -tionship between Service virtualisation (service clouds) and federation 17 are the

support of a new world of solutions defining the Future Internet Next generation networks and services 3 4 24 can not be conceived without

Management refers to the ability of such systems not just to configure, monitor and control a network element or service,

resources, the software that manages them, and the actors who direct such manage -ment. In federation management end-to-end communication services involve config

distributed software components that communicate across different networks 27 to satisfy end-to-end services client requests.

middleware and hardware levels (3. Analysis) that can be gathered, processed, aggre -gated and correlated (4. Mapping) to provide knowledge that will support management

Typical large enterprise systems contain thousands of physically distributed software components that communicate across different networks to satisfy client requests

that will enable enterprise application management systems to reconfigure software components to better adapt applications to prevailing network conditions.

7 Summary and Outlook In the future Internet new designs ideas of Federated Management in Future Internet

visualize software components, functionality and performance •Techniques for analysis, filtering, detection and comprehension of monitoring data

reconfigure or redeploy software components realizing autonomic application functionality •Guidelines and exemplars for the exchange of relevant knowledge between net

Computer Communications (July 2010), 63 pp http://www1. cse. wustl. edu/jain/papers/ftp/i3survey. pdf

Software systems for an Autonomic Internet. In: IEEE Globecom 2010, Miami, USA, 6 -10 december (2010 14.

Computer Society Press, Los Alamitos (2009 19. Brennan, R.,Feeney, K.,Keeney, J.,O†Sullivan, D.,Fleck II, J.,Foley, S.,van der Meer

IOS Press, Am -sterdam (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

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 under

•An identification of a core set of functions and underlying information models operations and interactions that these architecture have in common

multiple dimensions, encompassing sensor networks able to provide real world intel -ligence or the goal-oriented autonomous collaboration of distributed objects via local

-teraction capabilities to machines, software artifacts and humans connected to it The RWI assumes that the information flow to

c) Resource Users which represent the physical people or application software that intends to interact with Resources and Eoi

At its core, the proposed architectural model defines a set of entities and their rela

resources and the software components implementing the interaction endpoints from the user perspective (Resource End point †REP). Furthermore, the model distin

A REP is a software component that represents an interaction end-point for a physical resource.

is a device that executes the software process representing the REP As mentioned before, the resources

host itself, for example, a computer in the network or an embedded server may act as the REP host for a resource,

which is physically hosted on a sensor node connected to it This distinction is important when mobility, disconnections and other system dynamics

-sor nodes, from attacks by hosting their REPS on more powerful hardware Unlike other models, the Smart Object model considers also real-world entities in

The system is based on the OSGI service middleware and consists of two main sub systems: the service platform openaal and the

The PECES architecture PECES provides a comprehensive software layer to enable the seamless cooperation of embedded devices across various smart spaces on a

The PECES middleware architecture enables dynamic group-based communication between PECES applications (Resources) by utilizing contextual information based on a flexible context ontology.

-sources 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

PC IS w hi ch st or es W H A T W H ER

ASPIRE Advanced Sensors and lightweight Programmable middleware for Innovative RFID Enterprise applications, FP7, http://www. fp7-aspire. eu

/PECES PERVASIVE Computing in Embedded systems, FP7 http://www. ict-peces. eu /Semsorgrid4env Semantic Sensor Grids for Rapid Application Development for Environ

identify the resources within IDN-middleware independent of their physical locations in the lower layer are used Uniform Resource Locators (URL) to identify resource

The implementations of IDN-SA are a set of different software modules, one mod -ule for each layer.

Each module, implemented using an HTTP server, will offers a REST interface. The interaction between IDN-compliant applications and IDN-SA

IOS Press, Amsterdam (2010 2. Ayers, D.:From here to There. IEEE Internet Comput 11 (1), 85†89 (2007

and Encapsulation on Software Development Cost and Quality. IEEE Trans. Softw Eng. 21 (3), 200†208 (1995

1 University of Rome â€oela Sapienzaâ€, Computer and System Sciences Department Via Ariosto 25,00185 Rome, Italy

Future Internet architecture, Cognitive networks, Virtualization, In -teroperation 1 Introduction Already in 2005, there was the feeling that the architecture and protocols of the Inter

growth of small and/or mobile devices and sensors, of services and of security re -quirements began to show that current Internet is becoming itself a bottleneck.

-sources, computing resources, device characteristics) via virtualization and data min -ing functionalities; the metadata produced in this way are then input of intelligent

middleware functionalities, storage, computational, connectivity and networking capabilities, etc •Applications are utilized by the Actors to fulfill their requirements and needs ex

-tive 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 manage -ment functionalities. They interact with Actors, Resources and Applications through

Semantic Virtualization Enablers The Semantic Virtualization Enablers are in charge of virtualizing the heterogene -ous 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, interop

-tic Virtualization Enablers, take consistent control and management decisions concern -ing the best way to exploit

-dled by the Semantic Virtualization Enablers, in order to be actuated involving the proper Resources, Applications and Actors

Semantic Virtualization Enablers Cognitive Enablers Identity, Privacy Confidentiality Preferences, Profiling, Context Multimedia Content Analysis and Delivery

-minals, Base Stations, Backhaul network entities, Core network entities. The selec -tion and the mapping of the Cognitive Framework functionalities in the network enti

-ized through the implementation of appropriate Cognitive Middleware-based Agents in the following referred to as Cognitive Managers)

-haul Network entities, Core Network entities. There not exist a unique mapping be -tween the proposed conceptual framework over an existing telecommunication net

Indeed the software nature of the Cognitive Manager allows a transparent integration in the network nodes.

in particular, the Cognitive Enablers which are the core of the proposed architecture are independent of the peculiarities of the surrounding Resources, Actors and Appli

-tionalities are embedded in the Semantic Virtualization Enablers, while the Elabora -tion functionalities are embedded in the Cognitive Enablers.

5) The transparency and the middleware (firmware based) nature of the proposed Cognitive Manger architecture makes relatively easy its embedding in any

network entities, Core network entities: the most appropriate network entities for hosting the Cognitive Managers have to be selected environment by envi

the Cognitive Enabler software) can be added/upgraded/deleted through remote wired and/or wireless) control

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.

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 â€oecross-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

their virtualization, obtained thanks to the introduction of Semantic Virtualization En -ablers. 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 net

-ture Internet and ontology in computer systems. Section 2 describes the concepts of the Entity Title Model and the ontology at network layers.

application layer, without extending to the middle and lower layers of computer networks. In this research area, this paper aims to contribute to advancing the

Also can be created other kinds of classiï cation, such as hardware software and network, among others.

Some one of them (not all) can be used as resources in others relevant literature As the root superclass of one ontology is â€oeentity†or â€oething†the Entity

and translate them into functionality in computer networks Link Layer: This is the layer that has the responsibility to establish the link

-tation facility in software and hardware. However, in a collaborative eï €ort to others Future Internet works, the Entity Title Model has better contributions

layers running in a Distributed Operating system, without the traditional sock -ets used in TCP IP, is used the Raw Socket to enable the communication 19

use of the Raw Socket to communicate with the Distributed Operating system without the use of IP, TCP, UDP and SCTP

International Journal of Human and Computer Studies, 43 (5†6: 907†928 (1995 16 ITU-T:

Proceedings of the Fall Joint Computer Conference. AFIPS November 14-16, Volume 31, pp. 525†534.

de Gereë ncia e Operacâ¸aëoeo de Redes e Servicâ¸os-WGRS-SBRC (2009 28 Rubio-Loyola, J.,Serrat, J.,Astorga, A.,Chai, W. K.,Galis, A.,Clayman, S

IOS Press, Amsterdam (2009 32 Tselentis, G.,et al.:Towards the Future Internet-Emerging Trends from Euro

IOS Press, Amsterdam (2010 33 Tsiatsis, V.,Gluhak, A.,Bauge, T.,Montagut, F.,Bernat, J.,Bauer, M.,Villa

IOS Press, Ams -terdam (2010 34 Vissers, C.,Logrippo, L.:The Importance of the Service Concept in the Design

An Ontological Approach to Computer system Security. Information security Journal: A Global Perspective (2010 36 Wong, W.:

decreased traffic volume from its own content servers and reduced load of the servers as well as an improved performance of the application,

Peers and the ETMS servers, providing rating information, are located in these stub-ASES, which are interconnected via a hub-AS containing the

19th IEEE International Conference on Computer Communications and Net -works (ICCCN 2010), ZÃ rich, Switzerland (August 2010

software has already been developed for the initial scenario and it is simply a matter of deploying and adopting it on a wider scale.

use between two servers amongst the many possibilities-MPTCP naturally spreads traffic over the available paths

However, the protocol implementation should not impact hardware offloading of segmentation and check-summing. One reason that MPTCP uses

the two servers (at the bottom) travels over two paths through the switching fabric of the data

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 ac

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.

to think about the benefits and costs for OS vendors, end users, applications and ISPS (Internet service providers. Here also we see the importance of network ef

server sends â€oepremium†packets (perhaps for IPTV) as Conex-Not-Marked or Conex -Re-echo. Conex traffic is prioritised by the operator (â€oepremium serviceâ€.

and then the host†s software would automatically send the user†s premium traffic (Voip say) as Conex-enabled.

One important development in telecoms is virtualisation. Although the basic idea is longstanding, it has recently come to much greater practical importance with the rise

•Roll out of the software should be cheaper, therefore the expected benefits of the deployment can be less

immediately use the new (virtualised) software, so effectively a large number of users can be enabled simultaneously

â€oeroll back†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

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.

ACM SIGCOMM Computer Communications Review 40 (2)( 2010 8. Kostopoulos, A.,Warma, H.,Leva, T.,Heinrich, B.,Ford, A.,Eggert, L.:

Computer Communication Review 35,2 (2005 20. Key, P.,Massoulie, P.,Towsley, D.:Combined Multipath Routing and Congestion Con

microsoft. com/pubs/70208/tr-2005-111. pdf 21. Honda, M.:Call for contribution to middlebox survey (2010), http://www. ietf

-rors, draft-ford-http-multi-server, work in progress (2009 25. Raiciu, C.,Plunkte, C.,Barre, S.,Greenhalgh, A.,Wishcik, D.,Handley, M.:

access to the Internet via mobile devices, an ever-growing number of broadband users worldwide, lower entry barriers for nontechnical users to become content and ser

-age capabilities of servers and networking infrastructure. For example, routing table memory of core Internet routers can be considered a â€oepublic good†that retail ISPS

have an incentive to over-consume by performing prefix de-aggregation with Border Gateway Protocol (BGP.

when multiple candidate servers are available, a consumer may prefer the one offering better Qos,

the server that minimizes its cost; e g.,, this is possible if the provider operates a local

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.

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. The recognition that technology-in use frequently differs from technology -158 C. Kalogiros et al during-design is growing.

-cal solutions that achieve efficient spectrum usage for mobile devices. Following the increasing consensus on benefits of incorporating economic incentive mechanisms in

Internet-Emerging Trends from European Research, IOS Press, Amsterdam (2010 16. Trilogy: D10-Initial Evaluation of Social and Commercial Control Progress, 2009, Public

and provide an outlook to their mitigation, embedded in a system -atic security risk management process.

For example, it must be assumed that the core routers forward packets at line-speeds of tens of Gigabits per second, which requires ex

such as software components, individual nodes, or ASES. An upgraph of a node is the set of potential resources, that can be represented as a network map of domains and their

hardware accelerator has been developed for PLA 24 accelerating cryptographic operations Security Design for an Inter-Domain Publish/Subscribe Architecture 171

-dezvous core and scope-specific implementations of scope home nodes that imple -ment the functionality for a set of scopes

-ity implemented by the rendezvous core. We refer to our work in 5 for a detailed

be found in the rendezvous core, the subscription reaches the scope, which can then dynamically generate the response

-tions on Computer systems 2 (4), 277†288 (1984 14. Lagutin, D.,Visala, K.,Tarkoma, S.:

IEEE Computer Society Press, Los Alamitos (2004 20. Carpenter, B.:rfc1958: Architectural Principles of the Internet.

European Conference on Computer Net -work Defence, EC2ND (2009 23. Miller, V. S.:Use of elliptic curves in cryptography.

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.

in the scope of secure service engineering into comprehensive software and service life cycle support. Such a life cycle support must deliver

the present know how on building secure software services and systems: more stakeholders with diï €erent trust levels are involved in a typical service com

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 eï €ectively with this increased challenge is pertinent and well recognized by the research community and by the industrial one.

-cesses and tools for secure software development. This typically covers require -ments engineering, architecture creation, design and implementation techniques

developed software is secure. Assurance must be based on justiï able evidence and the whole process designed for assurance.

integrating the former results in (5) a risk-aware and cost-aware software devel -opment life-cycle (SDLC),

software development: from requirements over architecture and design to the composition and/or programming of working solutions.

-age existing mechanisms and methodologies for software construction as well as researching about new ways to take this complexity into account in a holistic

software service and/or system is a timely moment to enforce and reason about these security mechanisms,

The software architecture encompasses the more relevant elements of the ap -plication, providing either a static or/and a dynamic view of the application.

which comprise software elements, the ex -ternally visible properties of those elements, and the relationships among themâ€

early phase in the software development cycle The research topics one must focus on in this subarea relate to model-driven

Until this point in the software and service development process, diï €erent concerns †security among them †of the whole application have been sepa

-curring problem in software design. Design patterns, once identiï ed, allow reuse of design solutions that have proved to be eï €ective in the past, reducing costs

security-critical software systems 4 Security Support in Programming Environments Security Support in Programming Environments is not new;

-ing 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

technologies will be used to develop the actual software and systems The search for security support in programming environments has to take

-ciï c elements of the software design; other may simply be high priority security requirements that have articulated †such as the appropriate treatment of con

Middleware Aspects. The research community should re-investigate ser -vice-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

-gram to well-known programming principles or best-practices in secure software development. Emphasis will be put on language extensions that guarantee adher

run-time execution monitors that can provably enforce advanced security policies 19,3 including ï ned-grained access control policies usage control policies and

central role in the development of software based services to provide conï dence about the desired security level.

giving feedback at each stage of the software life cycle by checking that the related models and artefacts satisfy their functional and security requirements

enables to manage assurance throughout the software and service development life cycle (SDLC. The next section clariï es these issues

The main objective is to enable assurance in the development of software based services to ensure conï dence about their trustworthiness.

Our core goal is to incept a transverse methodology that enables to manage assurance throughout

the software development life cycle (SDLC. The methodology is based on two strands: A ï rst sub-domain covers early assurance at the level of requirements

of the software life cycle, automated generation in XML-based input data to maximize the eï ciency in the security testing process,

While the software development proceeds through incremental phases, the risk and cost analysis will undergo new iterations for

-gate through the software development phases and become more reï ned. In order to support the propagation of analysis results through the phases of the SDLC

In order to accommodate to a modular software development pro -cess, as well as eï €ectively handling the heterogeneous and compositional nature

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.:

agent-oriented software development methodology. Autonomous Agents and Multi -Agent Systems 8, 203†236 (2004

IEEE Computer Society Press, Los Alamitos (1981 doi: 10.1109/SFCS. 1981.32 10. Erlingsson, U.,Schneider, F. B.:

IEEE Computer Society Press, Los Alamitos (2000 11. France, R.,Fleurey, F.,Reddy, R.,Baudry, B.,Ghosh, S.:

IEEE Computer Society Press, Los Alamitos (2007 12. Giorgini, P.,Mouratidis, H.,Zannone, N.:Modelling security and trust with secure

SPAQU€ 08 (Int. Workshop on Software Patterns and Quality)( 2008 18. Lazouski, A.,Martinelli, F.,Mori, P.:

The vision of the Internet of Services (Ios) entails a major paradigm shift in the way ICT systems and applications are designed,

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 secu

-sumption time, thereby signiï cantly improving the all-round security of the Ios In this chapter, we give a brief overview of the main scientiï c and industrial chal

thus increase the development and public acceptance of the Ios We proceed as follows. In Sections 2 and 3, we discuss, respectively, some of

For instance, Tulafale 6, a tool by Microsoft Research based on Proverif 7, exploits abstract interpretation for veriï cation of web services

layer of software modules that carry out the translation from application-level speciï cation languages (such as BPMN and BPEL,

Google and the US Computer Emergency Readiness Team (US-CERT) were informed and the vulnerability was kept con

server to impersonate a user on any Google application. In 2, solutions that can be used to mitigate

, 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

with easy-to-use GUIS and translators to and from the core formal models should be devised and migrated to the selected development environments

out by building contacts with core business units. First, in the trail of the suc

6 Conclusions and Outlook As exempliï ed by these case studies and success stories, formal validation tech

-nologies can have a decisive impact for the trust and security of the Ios. The

advances will signiï cantly improve the all-round security of the Ios, and thus boost its development and public acceptance

IEEE Computer Society Press, Los Alamitos (2001 8. Bodei, C.,Buchholtz, M.,Degano, P.,Nielson, F.,Nielson, H r.:

Computer Society Press, Los Alamitos (2008 12. Ciobaë ca, S.,Cortier, V.:Protocol composition for arbitrary primitives.

IEEE Computer Society Press, Los Alamitos (2010 13. Clarke, E. M.,Grumberg, O.,Peled, D. A.:

Computer and Communications security (CCS 2010), pp. 351†360. ACM Press New york (2010 22. Moâ dersheim, S.,Vigano`,L.:

2007), http://docs. oasis-open. org/wsbpel/2. 0/OS/wsbpel-v2. 0-OS. pdf

IEEE Computer So -ciety Press, Los Alamitos (1977 26. T. Dierks and E. Rescorla. The Transport Layer Security (TLS) Protocol, Version

2 IBM Research †Zuâ rich, Ruâ schlikon, 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

from the convergence of computer server power, storage and network bandwidth It is a new business

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 across providers. Many of these developments can be expected to be transferred into

the Future Internet Core Platform project that will launch in 2011. This goes along with increased collaboration on open cloud standards under developments

Similarly, IBM has launched a FISMA compliant Federal Community Cloud in 2010 Other cloud providers also adapt basic service security to the needs of spe

Following its software-plus-services strategy an -nounced 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 oï €ers dedicated private cloud hosting and supports third-party or customer-site hosting. This allows tailor made solutions to speciï c security

the base security of Microsoft public cloud services is adapted to the targeted market. Whereas Microsoft uses, e g.,

, for the Oï ce 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

, Novell, IBM), virtual private network -ing (e g.,, Amazon Virtual Private cloud), encryption (e g.,, Amazon managed encryption services) and web traï c ï ltering services (e g.,

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 vir

-tual 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.

clouds are built often using low-cost commodity hardware that fails (relatively often. This leads to frequent failures of machines that may also include a subset

The consequence of these facts is automated that fault tolerance, problem -determination, 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

TCG) uses secure hardware to allow a stakeholder to perform attestation, i e.,, to obtain proof of the executables

Today, such channels are frozen often in hardware and thus cannot easily be reduced 218 R. Glott et al

5 Outlook †The Path Ahead Cloud computing is not new †it constitutes a new outsourcing delivery model

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

//www. symantec. com/connect/blogs/w32stuxnet-dossier 6. Chow, R.,Golle, P.,Jakobsson, M.,Shi, E.,Staddon, J.,Masuoka, R.,Molina

8. Computer and Communication Industry Association (CCIA: Cloud comput -ing (2009), http://www. ccianet. org/CCIA/files/cclibraryfiles/Filename

& service brokers (2010), http://www. processor. com/editorial/article. asp article=articles%2fp3203%2f39p03%2f39p03. asp

-ings of the 16th ACM conference on Computer and communications security Chicago, Illinois, USA. CCS †09, pp. 199†212.

of their data stored in a remote server or in the cloud. We introduce a

trusted event handler and a trusted obligation engine, which monitors and informs the user on the compliance with a previously agreed privacy

, servers services, applications) provided by the cloud, that are provisioned rapidly with a minimal management eï €ort and pay-peruse.

as setting and comparing user preferences with server privacy policies, ex -pressing conditions on complex secondary usage cases, specifying obligations

In fact, this sticky policy will be sent to the server and follow the data in all their lifecycle to specify the usage conditions

-ing 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

data before sending it to any server. For example, we can deï ne black or white lists for websites with whom we do not want to exchange our personal

it monitors all the events related to the usage of the collected data. These event notiï cations are handled by the obligation engine in order

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

should motivate the data collectors/processors to implement such technology Actually, in many cases, their business model relies on the as-less-restricted-as

In other words, we suppose that the server enforces correctly the sticky policies, but, actually, nothing prevents him from

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

-sibility 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

console could be structured, this monitoring console could of course be more complex. The remote monitoring console adds more transparency and more con

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 tam

authorization language (EPAL 1. 1). IBM Research Report (2003 Data Usage Control in the future Internet Cloud 231

Trust and tamper-proof software delivery. In: Proceed -ings of the 2006 international workshop on Software engineering for secure sys

It provides a core infrastructure, and also a playground for future discoveries and innovations, combin

-strates the practical usefulness of federation and virtualisation in heterogeneous testbeds These multipath routing slice experiments were performed over multiple federated

Resource Adapter (a concept similar to device drivers) wraps a domain†s resource API in order to create a homogeneous API defined by Panlab.

server, an application server, a database and a workload generator, which produces the appropriate requests.

using Xen server technology, which allows regulating system resources such as CPU usage and memory, and provides also a monitoring tool, Ganglia, that measures net

-work metrics, such as round trip time and other statistics, and resource usage in vir -tual machines

using Xen server technology to regulate CPU usage. During this scenario the adaptive admission control and resource allocation algorithm is tested against network metrics

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 pro

-gramming 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 genera

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.

â € Linux machines for the RUBIS based work load generators â € A Linux machine for the hosting the algorithm unit, capable of compiling C and

Java software â € 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 (i e. cannot login) so there is a need to encapsulate the monitoring and provisioning capabilities.

Virtual machines managed by a XEN server. The implemented RAS instantiate all these Virtual machines and configure the internal components according to end-user

memory, hard disk size, number of clients, ramp up time for the requests and a pa

-bed, memory, hard disk size, username, password and IP to connect to the RUBIS application resource.

cpu capacity of the resource A Use-Case on Testing Adaptive Admission Control 241 Fig. 2. The Resource adapters of the available testbed resources

displays the RADL definition for the RUBIS application server The Configuration Parameters section describes the exposed parameters to the end

Figure 4 displays the use case setup as can be done inside the VCT tool of Panlab.

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 in

listing displays how we can access the resources of this VCT. FCI creates a java class

back the CPU usage of the database resource 5 Conclusions The results of running an experiment in Panlab are encouraging in terms of moving

In such an environment federation and virtualization of resources are key features that should be supported in a future Internet.

Network Virtualization (NV) techniques 5, 17 allow the establishment of such separate slices on top of a joint physical infrastructure (substrate.

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 applica

to the high costs of such hardware. In short, as network scientists, we need larger testbeds in order to supplement theoretical analysis and validate theoretical re

-tion 7 provides a brief summary and outlook to the enhancements of federated facilities 2 Experiment Objectives and Requirements for a

reservation 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.

arbitrary software on the G-Lab nodes. We assume that such features are of interest for many experimenters,

-erators and developers to oï €er their measurement results and software tools to the public and to share their experience.

-passe 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.:

Network virtualization: Breaking the performance barrier. ACM Queue,(Jan./Feb. 2008 18. Santos, T.,Henke, C.,Schmoll, C.,Zseby.

Computer Communication Review 35 (5), 71†74 (2005 20. Phuoc Tran-Gia. G-Lab: A Future Generation Internet Research Platform (2008

Virtualization. In: Proc. of the 16. Kivs 2009, Kassel, Germany (Mar. 2009 24. Zinner, T.,Tutschku, K.,Nakao, A.,Tran-Gia, P.:

Transmission for Transport Virtualization: Analyzing Path Selection. In: Proceed -ings of the 22nd International Teletraï c Congress (ITC), Amsterdam, Netherlands

-tures by using the Self-NET software for self management over a Wimax network environment. The monitoring and configuration capabilities that differ

the so called closed control loop or Monitor-Decide-Execute Cycle (MDE) and con -sists of the Network Element Cognitive Manager (NECM) and the Network Domain

which is a software tool that generates traffic at both Uoa end machines. This is a

Fig. 1. Octopus testbed Wimax and Self-NET software federation ITGLOG), printing and plotting specific metrics (ITGDEC, ITGPLOT) and remotely

software and deployment of IP routing and tunneling between Octopus and Self-NET environments 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) set

The NECM of the Wimax BS constantly monitors network device statistics (e g UL/DL used capacity, TCP/UDP parameters, service flows), which are periodically

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

the future innovation of both the networking technology that lies at the Internet†s core

providing a natural complement to the virtualization of resources -by setting up and tearing down composed services, based on negotiated SLAS.

of the current network typically resides in client stations and servers, which interact with network elements (NES) via protocols such as SNMP (Simple Network Man

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. In addition, security risks currently present in network environments request for immediate attention.

information and enabling almost real-time interaction between the virtual world and real world. In particular, autonomous self-organizing systems are beginning to emerge

Among the core drivers for the FI are increased reliability, enhanced services, more flexibility, and simplified opera

high quality, on a real-time basis. Self management can offer decentralized monitor -ing and proper decision-making techniques so that appropriate optimization hints can

Monitor Phase 2. 760 2. 561 3. 137 2. 547 Tim e se c Fig. 4. Channel Selection Duration

-ment, virtualization of systems and network resources, advanced and cognitive net -working of information objects), have to â€oere-define†the overall FI network architec

IOS Press, Amsterdam (2009 10. Organization for Economic Co-operation Development (OECD: The Seoul Declaration

IBM Corporation (2008 30. Prehofer, C.,Bettstetter, C.:Self-organization in Communication Networks: Principles and Design Paradigms.

-ence on Ultra Modern Telecommunications (ICUMT-2009), pp. 1†6. IEEE Computer So -ciety Press, Los Alamitos (2009

open-source multimedia framework, player and server http://www. videolan. org/vlc J. Domingue et al. Eds.):

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

PDAS, laptops and other network-enabled devices. Thus, a fitness function is pre -sented which is able to evaluate the eligibility of each candidate node

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.

-tional Conference on Computer Communications (2006 3. Rong, B.,Hafid, A.:A Distributed Relay Selection algorithm for Cooperative Multicast in

Computer Communications 31, 1763†1776 (2008 5. Verma, A.,Sawant, H.,Tan, J.:Selection and navigation of mobile sensor nodes using a

Computer Net -works 47, 445†487 (2005 9. Akyildiz, I.,Lee, W.,Chowdhury, K.:CRAHNS:

Future Internet, Virtualization, Dynamic Provisioning, Vir -tual Infrastructures, Convergence, Iaas, Optical Network, Cloud 1 Introduction

links, which in the core-network segment are mostly based on optical transmis -sion technology, but also in the access segments gradual migration to optical

-demand access to IT hardware or software resources over the Internet. Clouds are revolutionizing the IT world 11,

-sioning and the virtualization paradigm with dynamic network provisioning as a way towards such a sustainable future Internet.

-virtualization layer, the enhanced control plane, that corresponds to the network management layer, and the service middleware layer.

Each layer is responsible for implementing diï €erent 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

-acting with the virtualization layer 3. Finally, a service middleware layer is introduced to fully decouple the

physical infrastructure from the service level. It is an intermediate layer between applications running at the service consumer†s premises and the

-virtualization layer is implemented as the Logical Infrastructure Composition Layer (LICL) and the enhanced control plane as the NCP+.

SML Service Middleware Layer VI Virtual Infrastructure VIO Virtual Infrastructure Operator VIO-IT Virtual IT Infrastructure Operator

Enterprise Information system externally on a Cloud rented from a Software -as-a-Service (Saas) provider.

between the NCP+and the service middleware layer (SML) via a service -to-network interface, named NIPS UNI during the entire VI service life cycle

19 http://www. ens-lyon. fr/LIP/RESO/Software/vxdl/home. html 316 P. Vicat-Blanc et al

of amount of CPU), while the NCP+computes the most eï cient combination of end-points and network path to be used for the speciï c service.

the infrastructure-virtualization layer, in order to trigger the procedures for the Virtual Infrastructure†s dynamic replanning on the network side, besides the IT

-opment activities of the GEYSERS project to achieve the complete software stack and provide the proof of concept of these architectural considerations

of the diï €erent software components creating and managing optical virtual in -frastructures. The other goal is to evaluate the performance and functionality of

CCGRID †09, p. 1. IEEE Computer Society Press, Los Alamitos (2009 doi: 10.1109/CCGRID. 2009.97

A Novel Architecture for Virtualization and Co-Provisioning of Dynamic Optical Networks and IT Services.

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

software on-demand, from a shared pool, with minimal interaction or knowledge by users. Cloud services can be divided into three target audiences:

software developers and users as follows6 ⠀ Infrastructure as a service †offering resources such as a virtual machine or

⠀ Platform as a service †providing services for software vendors such as a soft -ware development platform or a hosting service

The ability to trade IT-services as an economic good is seen as a core feature of

1 Intel, Ireland, {joe. m. butler, michael. nolan}@ intel. com 2 Telefã nica Investigaciã n y Desarrollo, Spain, juanlr@tid. com

3 SAP AG, Germany, wolfgang. theilmann@sap. com 4 ENG, Italy, francesco. torelli@eng. com

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

-cally, a service is dependent on many other services, 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 manag -ing SLAS in the future Internet.

The framework supports the configuration of com -plex service hierarchies with arbitrary layers. This allows end to end management of

-ness, software, and infrastructure) on the other With a set of four complementary use case studies, we are able to evaluate our ap

2) supports arbitrary service types (business, software, infrastructure) and SLA terms,(3) covers the complete SLA and service lifecycle with consistent interlinking

business, software and infrastruc -ture. The framework communicates to external parties, namely customers who (want

On the highest level, we distinguish the Framework Core, Service Managers (infra -structure 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.

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

Software Service <<adjust >>deployed infrastructure service deployed software service <<negotiate >>framework core Fig. 1. Overview of the SLA Framework Reference Architecture

While all framework components come with default implementations they can also easily be extended or enhanced for more specific domain needs.

Similarly, the pro -vided meta-models come with clear extension mechanisms, e g. to specify additional service level terms

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.

We assume a virtualisation-enabled data centre style configura -tion of server capacity, and a broad range of services in terms of relative priority

resource requirement and longevity. As a support service in most enterprises, IT is expected to deliver application

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

a software package (an application) but also some business-level activities, such as a support contract. 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 mid -dleware components which are used equally for different applications.

At the lowest layer, there are the infrastructure resources, delivered through an internal or external cloud.

Application SLA is mainly about the throughput capacity of the software solution, its response time, and the provider internal costs required for the offering.

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, mem

-ory, 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

the different components, e g. the performance characteristics of the middleware. We collected a set of model-driven architecture artefacts, measurements,

The main aim of the Service Aggregation use case is the service-enabling of core

integrates software layer (from SLA@SOI framework architecture. And finally Bank prototype is implemented using the top layer, business.

wrappers by Software SLA Manager in an application server and also the provision of the infrastructure driven by Infrastructure SLA Manager (using the appropriate ser

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 Telefã nica 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.

For example, while typical software/hardware guarantee terms constraint the quality of each single execution of a service, in this use case the guarantee terms constraint

-purpose SLA management framework that can become a core element for managing SLAS in the future Internet.

network, software, and business artefacts. Four complementary industrial use cases demonstrated the applicability and relevance of the approach.

-opment activities via an Open source Project. The first framework version fully pub -lished 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, and reproduction

Whitepaper IBM developerworks (March 2008), http://www. ibm. com/developer works/autonomic/library/ac-edge4

5. SLA@SOI Open source Framework. First full release by December 2010 http://sourceforge. net/projects/sla-at-soi

Management and Virtualisation planes), presented in 3, to better monitor the networks, as the semantic information can directly be handled by the Net

there are some limitations for the software-driven control network infrastruc -ture, formed by the OSKMV,

contribute to the translations of the MBT (Model-based Translator) software package, by the use of the FINLAN formal representation in OWL.

use of the CPU memory assignment, packets lost and others. The invocation of the methods can

The Autoi open source implements a scal -able and modular architecture to the deployment, control and management of

the Session Broker and the Virtualisation Broker. There are many others but these are essentials. As the communication between the Autoi modules is done

-sites for an instance (memory size, storage pool size, number of virtual CPUS 346 E. Santos et al

{private Serversocket server private int port=43702 private Socket s=null public static Knowledgeplane KP=null

+""s1=server. accept With the use of the FINLAN library this communication can be done replacing

the FINLAN ontology in Linux operating system kernel level, allowing the facil -ities in its use in diï €erent programming languages,

would be available at the operating system level 4 Conclusions This paper has presented the FINLAN ontology works in a collaboration per

Future work will implement the FINLAN ontology at the Linux kernel level and run performance and scalability experiments with diï €erent Future Internet

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

Integracâ¸aëoeo de Servicâ¸os em Ambientes Heterogeë neos: uso de Semaë ntica para Comunicacâ¸aëoeo Entre Entidades em Mudancâ¸as de Contexto.

The Internet of Services is seen as a core com -ponent of the Future Internet â€oethe Future Internet is polymorphic infrastructure, where the bounda

consuming of functionalities of existing pieces of software. In particular, WSDL is used to provide structured descriptions for services, operations and endpoints, while

The use of services as the core abstraction for constructing Linked Data applications is therefore more generally applicable than that

users ratings, tags and comments about services in a separate server. On the basis of

software clients are able to access easily underlying descriptions Linking †linking between descriptions should be supported to facilitate the reuse

Architectural styles and the Design of Network-based Software Architec -tures. Phd Thesis, University of California (2000

core parts, with Qos assurance is seen. A flexible way of usage †based on virtual -ised overlays †can offer a strong support for the transportation of multimedia

•Third, the todays†software technologies support the creation and composition of services while being able to take into account information regarding the trans

Based on virtualization, the network can offer enhanced transport and adaptation-capable services This chapter will introduce

The virtualisation as a powerful tool to overcome the Internet ossification by creating overlays is discussed in 10-11.

single or multiple core network domains having content aware processing capabilities in terms of Qos, monitoring, media flow adaptation, routing/forwarding and security

radical approach can also be envisaged towards full virtualization (i e. independent management and control per VCAN

At its turn, the SM@SP instructs the SP/CP servers how to mark the data packets

server AS3 VCAN1/MQC1 VCAN2/MQC2 VCAN3/MQC3 L2, L3, L4 headers High level headers

avoiding per-flow signalling in the core part of the network. In the new architecture MANE also can act as content â€oecachesâ€,

-tween CPS and NPS for hosting or co-locating CP€ s content servers in NPS€ premises

IOS Press, Amsterdam (2009 4. Schã nwã¤lder, J.,et al.:Future Internet=Content+Services+Management.

IOS Press, Amsterdam (2009 6. Huszã¡k, Ã.,Imre, S.:Content-aware Interface Selection Method for Multi-Path Video

IOS Press, Amsterdam (2009 8. Martini, M. G.,et al.:Content Adaptive Network Aware Joint Optimization of Wireless

IOS Press, Amsterdam (2009 10. Anderson, T.,et al.:Overcoming the Internet Impasse through Virtualization. Com

-puter 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:

video streaming, due to a bandwidth bottleneck at the server side from which all users request the content.

MDC combined with path/server diversity offers robust video delivery over unreliable networks and/or in peer-to-peer

-dence of software proposal for Wavelet Video Coding Exploration group, ISO/IEC JTC1/SC29/WG11/MPEG2006/M13146, 76th MPEG Meeting, Montreux, Switzerland

search engines are expected to be able to understand underlying semantics in content and match it to the query.

-ios as semantic image retrieval. However, the semantic gap lying between levels 1 and 2 hampers the progress of multimedia retrieval area 19.

took only a few seconds on a PC with Pentium D CPU 3. 40ghz and 2. 00gb of

IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp. 994†999 (1997 4. Chang, E.,Goh, K.,Sychay, G.,Wu, G.:

IEEE Computer Society Conference on Computer Vi -sion and Pattern Recognition, vol. 2 (2003 12.

IEEE Computer Society Con -ference on Computer Vision and Pattern Recognition, vol. 2 (2004 13.

Kherï, M. L.,Ziou, D.:Image collection organization and its application to in -dexing, browsing, summarization,

Signal Processing ICASSP€ 04, vol. 3, IEEE Computer Society Press, Los Alamitos 2004 16. Naphade, M. R.,Huang, T s.:

Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, vol. 1, pp. 200†206 (1999

Innovation lies at the core of smart enterprises and includes not only products, services and processes but also the

the federation and integration of appropriate software building blocks. A new genera -tion of enterprise systems comprising applications

of utility-like software services High-value Future Internet applications are also foreseen in the domain of living

In recent years, software development methods and technologies have markedly evolved, with the advent of SOA 15, MDA 16, Ontologies and Semantic web, to

and realising enterprises software applications In essence, while enterprise management and planning services will be increas -ingly available from the †cloudâ€, in a commoditised form, the future business needs

the Internet of Services (Ios), Internet of things (Iot) and smart ob -jects, Internet of Knowledge (Iok), Internet of People (Iop.

services, tools, software packages, interfaces and user interaction solutions that are not available at the present time.

of software architectures. This will be possible if such software architectures will correspond to the enterprise architectures,

and will be composed by elements tightly coupled with business entities. The achievement of this objective relies on a number

based on the notion of a FINER, seen as the new frontier to software components aimed at achieving agile system architectures.

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

) The basic idea of software componentization is quite the same as software modularization, but mainly focused on reuse.

that in a short time imposed itself over the preexisting modular software develop -ment techniques.

OOP aims at developing applications and software systems that provide a high level of data abstraction and modularity (using technologies such as

Agents mark a fundamental difference from conventional software mod -ules in that they are inherently autonomous and endowed with advanced communica

highly interoperable software architectures through the spread of the Service-Oriented approach, and the consequent proliferation of Service-Oriented Architectures (SOA

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.

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.

manage a new generation enterprise software architectures. Cloud computing repre -sents an innovative way to architect and remotely manage computing resources:

over the Internet and the hardware and system software in the datacenters that provide those services 12.

approach, to ease software development processes. Conversely, we propose to base a FINES architecture on building blocks based on business components.

or responding (as server) to request messages. It is structured according to the grounding of OWL-S

•Tangible entity, from computers to aircrafts, to buildings and furniture •Intangible entity, for which a digital image is mandatory

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

Component-oriented software development, Spe -cial issue on alaysis and modeling in software development, pp. 160†165 (1992

11. Petritsch, H.:Service-Oriented Architecture (SOA) vs. Component Based Architecture white paper, TU Wien (2006), http://whitepapers. techrepublic. com. com

IOS Press, Amsterdam (2010 16. Papazoglou, M. P.:Web Services: Principles and Technology. Prentice-hall, Englewood

-cused on microprocessor design, computer design, power-on-demand architectures and virtual machine consolidation techniques. However, a micro-level energy effi

Large ICT companies, like Microsoft which consumes up to 27mw of energy at any given time 1,

and in the server farms is not considered since no special equipment is deployed in the GSN

will be developed to leverage virtualization, which helps to migrate virtual infrastruc -ture resources from one site to another based on power availability.

Core nodes are linked by an underlying high speed optical network having up to 1, 000 Gbit/s bandwidth capacity provided by CANARIE.

a convergence of server and network virtualizations as virtual infrastructure manage -ment. The GSN as a network architecture is built with multiple layers, resulting in a

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

robust software while running trials on a range of network equipment The Mantychore FP7 project allows the NRENS to provide a complete, flexible

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 (Canada GSN-Calgary (Canada

Fig. 2. Architecture of green nodes (hydro, wind and solar types 424 K. K. Nguyen et al

by green energy and adjust the network to the needs controlled by software. The cost

such as routers and servers, is not considered, 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

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

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

then pushes Virtual machines (VMS) or software virtual routers from the hub to a sun or wind node (spoke node) when power is available.

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 capabili -ties 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

servers with a high density connection local network. The migration involves four steps i) Setting up a new environment (i e.,

servers and network devices in a multi-domain environment Fig. 4. Iaas Framework Architecture Overview

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,

converges server and network virtualizations. Whilst most of cloud management solu -tions in the market focus particularly on computing resources, Iaas Framework compo

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, launch, monitoring,

and interoperate with software outside of the GSN Along with the participation of international nodes, there is an increasing need of

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 availabil -ity. There is a domain controller within each data center or a set of data centers shar

Virtualization techniques are shown to be the most appropriate solution to manage such a network and to migrate data cen

-passing peripheral and less developed cities. It also emphasises the process of eco -nomic recovery for welfare and well-being purposes.

-driven innovation ecosystems may evolve to constitute the core of â€oe4p†(Public -Private-People-Partnership) ecosystems providing opportunities to citizens and busi

-tion 4. Finally, section 5 presents conclusions and an outlook 2 City and Urban Development Challenges

such as mobile devices (e g. smart phones), the semantic web, cloud computing, and the Internet of things (Iot) promot -ing real world user interfaces

-ing, wireless networks, middleware and agent technologies as they become embedded into the physical spaces of cities.

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, dis -tributed intelligence, crowdsourcing, online collaboration, broadband for innovation, social

Smart cities, from smart phones, mobile devices, sensors, embedded systems, smart envi -ronments, smart meters, and instrumentation sustaining the intelligence of cities

-connection of mobile devices, sensors and actuators allowing real-world urban data to be collected and analysed, will improve the ability to forecast

cities with embedded systems, smart devices, sensors, and actuators, offering real -time data management, alerts, and information processing,

in parallel with the Internet of things and embedded systems, providing new oppor -tunities for content management 12,13. Media Internet technologies are at the

software components City-based clouds Open and federated content platforms Cloud-based fully connected city

Software agents and advanced sensor fusion; telepresence Demand for e-services in the domains outlined in Fig. 1 is increasing,

There is a critical gap between software applications and the provi -sion of e-services in terms of sustainability and financial viability.

Open source communities may also sub -stantially 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.

At the core of Periphã ria lies the role of Living Labs in constituting a bridge

appreciation of cultural heritage) with networks of video-cameras used to monitor public spaces. In addition, the integration of these services occurs in the Living Lab

5 Conclusions and Outlook In this paper we explored the concept of â€oesmart cities†as environments of open and

IBM Journal of Research & development 53 (3), 338†353 (2009 11. European commission: Growing Regions, Growing Europe:

Things (Iot) and Internet of Services (Ios), can become building blocks to pro -gress towards a unified urban-scale ICT platform transforming a Smart City

and at the service level (Ios as a suit of open and standardized en -ablers to facilitate the composition of interoperable smart city services.

design of urban-scale ICT platforms, three main core functionalities can be identified •Urban Communications Abstraction.

•The Internet of Services (Ios: flexible, open and standardized enablers that facili -tate the harmonization of various applications into interoperable services as well as

-nents of the Future Internet, namely Iot and Ios, can be essential building blocks in future Smart Cities open innovation platforms.

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

smart meters and monitoring devices can help monitor and manage water consumption, heating, air-conditioning, lighting and physical secu

Ios evolution must be correlated undoubtedly with Iot advances. Otherwise, a number of future Smart City services will never have an opportunity to be con

-ing Iot and Ios at the city scale Starting with the benefits of Iot technologies, they are twofold:

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 innova

-tion platform through the proper harmonization of Ios and Iot. There can be a long

Thus the integration of innovative principles and philosophy of Ios will engage collective end-user intelligence from Web 2. 0

The technological challenge of developing the Ios has been assumed at EU level, and actions are being initiated to overcome the undesirable dissociation between technologi

Ios resources Testbed 1 USN-Enabler Service 1 Adaptation & Homogeneization Testbed Control Layer GSDP

Ios federation level Iot federation level NGN /Telco2. 0 Web2. 0 Service 2 Iot Service Service n

Fig. 1. Global Service Delivery Platform (GSDP) integrating Iot/Ios building blocks 3 Developing Urban Iot Platforms

This capability will allow a seamless link between Iot and Ios, as discussed in Section 2

Iot Core Platform Working group discussion 25, multiple different approaches for First Generation Iot-platforms are currently being implemented.

sensors networks deployed throughout the city and the core IP communication network), the main responsibility is to provide independence from the communica

needed to enable an evolving FI based on the Iot and Ios paradigms Functionalities required to support services are offered both in synchronous and

Opencom Middleware Mobility support Horizontal support Federation support S ecurity, Privacy and Trust Fig. 3. Smartsantander:

•Tourism information in different parts of the city through mobile devices using visual and interactive experiences and in different languages

initiatives on both Iot and Ios areas as WISEBED 25, SENSEI 8 and the USN

the subsystems (collectively, the Smartsantander middleware) that provide the func -tionality described by these requirements

Future Internet potential, through Iot and Ios, for creating new real-life applications and services is huge in the smart city context.

-twining Iot and Ios worlds. Referred Iot USN platform is currently being evolved with the addition of new capabilities,

Research Challenges for the Core Platform for the Future Internet. In: M Boniface, M. Surridge, C. U (Eds.

/activities/foi/library/docs/fippp-research-challenges-for-core -platform-issue-1-1. pdf 5. Sundmaeker, H.,Guillemin, P.,Friess, P.,Woelfflã, S. eds.:

) Towards the Future Internet, IOS Press, Amsterdam (2009 17. Fisher, S.:Towards an Open Federation Alliance.

Summary and Outlook References An Architectural Blueprint for a Real-world Internet Introduction The Real world Internet

The Need for Engineering Secure Software Services Research Focus on Developing Secure FI Services Security Requirements Engineering

Conclusions and Outlook Trustworthy Clouds Underpinning the Future Internet Cloud computing and the Future Internet Trust and Security Limitations of Global Cloud Infrastructures

Outlook †The Path Ahead Data Usage Control in the future Internet Cloud Introduction Primelife Privacy Framework

Conclusions and Outlook References Smart Cities at the Forefront of the Future Internet Introduction Iot and Ios as ICT Building blocks for Smart Cities

Developing Urban Iot Platforms The Need of Urban Scale Experimental Facilities Conclusions References Author Index

>/NLD (Gebruik deze instellingen om Adobe PDF-documenten te maken die zijn geoptimaliseerd voor weergave op een beeldscherm, e-mail en internet.

De gemaakte PDF-documenten kunnen worden geopend met Acrobat en Adobe Reader 5. 0 en hoger

>/ENU (Use these settings to create Adobe PDF documents best suited for on-screen display, e-mail,

Created PDF documents can be opened with Acrobat and Adobe Reader 5. 0 and later /DEU<FEFF004A006F0062006F007000740069006F006E007300200066006F00720020004100630072006 006200610074002000440069007300740069006c006c0065007200200037000d00500072006f006 00750063006500730020005000440046002000660069006c0065007300200077006800690063006 00200061007200650020007500730065006400200066006f00720020006f006e006c0069006e006 002e000d0028006300290020003200300031003000200053007000720069006e006700650072002 005600650072006c0061006700200047006d006200480020

Adobe Common 1. 0 /Othernamespaces <</Asreaderspreads false /Cropimagestoframes true /Errorcontrol/Warnandcontinue /Flattenerignorespreadoverrides false /Includeguidesgrids false

Adobe Indesign 4. 0 /Omitplacedbitmaps false /Omitplacedeps false /Omitplacedpdf false /Simulateoverprint/Legacy >><Addbleedmarks false /Addcolorbars false

/Addcropmarks false /Addpageinfo false /Addregmarks false /Convertcolors/Converttorgb /Destinationprofilename (srgb IEC61966-2. 1 /Destinationprofileselector/Usename

Adobe Creativesuite 2. 0 /PDFXOUTPUTINTENTPROFILESELECTOR/NA /Preserveediting false /Untaggedcmykhandling/Usedocumentprofile /Untaggedrgbhandling/Usedocumentprofile /Usedocumentbleed false >>>setdistillerparams

< Back - Next >

Overtext Web Module V3.0 Alpha
Copyright Semantic-Knowledge, 1994-2011