CELTIC Project Description
CELTIC Project Description (PD)
Project Identification
Project acronym: / MARCHProject name: / Multilink architecture for multiplay services
Project number / CP5-013
Project Coordinator (name): / Terje Tjelta
Email: /
Company / Telenor ASA
Telephone: / +47 90786424
Key Information
Issue date: Version 3 / 10 March 2009Kick-off meeting (date): / 17-18 June 2008
Start date of Project: (month/year): / June 2008
Provisional Date for the Mid Term Review (month/year): / 11/2009
Planned completion date (month/year): / May 2011
Duration (months): / 36
CELTIC label approved on (date): / 26 November 2007
Planned effort and budget / Total / 2008 / 2009 / 2010+2011
Effort in Person Years (PY) / 59.0 / 4.7 / 30.5 / 23.9
Total budget (kEUR) / 7246.2 / 707.3 / 3468.2 / 3070.6
Additional costs for travel, equipment, etc. (included in total) / 505.0 / 51.0 / 215.5 / 238.5
Completion of legal documents
Confirmation Letters signed by all partners / All partners signed except ThomsonProject Cooperation Agreement approved/ signed / Started the process
Declaration of Acceptance signed / Signed by all partners
Project Change Request issued (if any) / 9 June 2008: Issued to set the start of project to 1 June 2008
4 September 2008: The first PD issued from the FPP including split of Simula into Simula and Lividi.
19 September 2008: To include two new partners, WiNetworks and LiveU from Israel.
10 March 2009: Mainly to take account for Thomson, PLUS and eduhi leaving the project and Matchmind replacing Moviquity.
Project Abstract (Project overview - 25 lines max.):
The trend today is for easy, ubiquitous access to advanced broadband services. People will expect to be able to connect to a broadband network for business and personal use, irrespective of time, terminal and location. The consumed content may be generated in any part of the network and will include both professional production and user generated content. Access to content will involve both peer-to-peer usages as well as centralised service delivery.
MARCH aims at significantly lowering the cost of converged broadband service delivery by establishing scalable multilink network architectures. Although a multiplicity of fixed and mobile networks exists today there has been little real attempt to achieve harmonisation or interworking. Today’s terminals incorporate multiple access technologies, but the strategies for connecting to IP networks generally result in simple selection of one of the available networks. The potential for simultaneously using the available access technologies in a multi-link manner has yet to be exploited.
MARCH addresses key issues for ubiquitous broadband access with a particular focus on service provision under the new broadband reality and network architecture effective for connecting all types of users, either fixed or on the move. The project aims at convergence of broadband, mobile, and broadcast approaches and its application to urban and rural deployment.
Focus areas are:
1) Multilink network technology
Convergence between broadband, mobile, and broadcast services is central in delivering ubiquitous and seamless services. MARCH focuses on the techniques enabling a multilink architecture whereby terminals operate dynamically over several broadband/broadcast access networks simultaneously enabling intelligent multi-service provision offering enhanced service availability and improved resource utilisation.
2) Modular scalable architecture
MARCH will recommend a modular scalable architecture for multilink networks and provide standardisation inputs such that networks can be realised at significantly lower cost. Networks have to scale well both allowing initial development with low demand and subsequent flexible growth.
Consortium Overview
Company name(Leading contractor first) / Country / Type *) / Address
Telenor ASA [Telenor] / Norway / T / Snaroyvein 30, Fornebu
Budapest University of Technology and Economics [BME] / Hungary / U / Goldmann Gy. tér 3, Budapest
Bitnet CCSS [Bitnet] / Romania / S / Madach 6, Cluj-Napoca
Matchmind S.L. [Matchmind] / Spain / I / C/ Ombú 3, 10ª, Madrid
TELECOM CASTILLA-LA MANCHA, S.A. [TCLM] / Spain / T/S / Poligono, Santa María de Benquerencia, Toledo
TECHNICAL UNIVERSITY OF CLUJ-NAPOCA [TUCN] / Romania / U / 15, C. Daicoviciu Street, Cluj-Napoca
ReignSoft Kft [Reignsoft] / Hungary / S / Gesztenyés u 9 IX/3., Miskolc
SIMULA Innovation [Simula] / Norway / S / Martin Linges vei 17, Fornebu
Lividi / Norway / S / Lividi c/o SIMULA Innovation, Martin Linges vei 17, Fornebu
WiNetworks Ltd [WiNetworks] / Israel / S / WiNetworks Inc, 32 Maskit St , 3rd floor, PO Box 12412, Herzeliya
LiveU [LiveU] / Israel / S / LiveU Ltd., 9 Atir Yedda St., Kfar Saba
INDRA Sistemas S.A [INDRA] / Spain / I / Paseo Club deportivo 1. Parque empresarial LA FINCA 28223, Pozuelo de Alarcón (Madrid)
*) Company Type: I = Industry, T = TelCO, S = SME, R = Research, U = University, A = Administration (public)
Confidential 2 (63)
CELTIC Project Description
Table of Content
Project Identification 1
Key Information 1
Completion of legal documents 1
Project Abstract (Project overview - 25 lines max.): 2
Consortium Overview 2
Table of Content 4
1 Project Summary 5
1.1 Main focus 5
1.2 State-of-the-art situation 5
1.3 Relevance to market needs and expected impact 7
1.4 Technological innovation and strategic relevance 7
1.5 Relevance to CELTIC 8
1.6 Major results 9
1.7 Contribution to Standards 9
1.8 Dissemination activities and exploitation of results 10
2 Project and work organisation 11
2.1 Project structure 11
2.2 Project Calendar 21
2.3 Risk assessment 21
2.4 Work breakdown 22
2.4.1 WP 1 22
2.4.2 WP 2 24
2.4.3 WP 3 27
2.4.4 WP 4 30
2.4.5 WP 5 34
2.4.6 WP 6 36
2.4.7 WP 7 38
2.4.8 Deliverables Overview 43
3 Project management 46
3.1 Management structure and procedures 46
3.1.1 Management structure 46
4 Details on budget and funding 49
4.1 Budget and Funding Situation of Participants 49
4.2 Specification of additional budget (equipment, other costs) 49
5 Overview of the Consortium 50
5.1 Description of the consortium 50
5.1.1 Consortium overview 50
5.1.2 Consortium partners and key personnel 50
5.2 Contact details 61
5.3 Contact and funding information by Public Authorities 63
6 Effort and budget tables 65
6.1 Effort allocation per partner and year 65
6.2 Total budget allocations per partners 69
6.3 Total budget allocations per country 69
6.4 Budget for equipment and other costs 70
7 Appendices 71
References 71
1 Project Summary
1.1 Main focus
The trend today is for widespread, easy access to advanced broadband services. People will expect to be able to connect to a broadband network for business and personal use, irrespective of time, terminal and location. The consumed content may be generated in any part of the network and will include both professional production and user generated content. Access to content will involve both peer-to-peer usages as well as centralised service delivery.
MARCH aims at significantly lowering the cost of converged broadband service delivery by establishing scalable multilink network architectures, as illustrated in Figure 1. The broadcast and telecommunications landscape is today characterised by a multiplicity of heterogeneous access technologies, both fixed and mobile. Users can connect to a broadband service using one of several technologies; they have access to a mobile network and to a satellite network and more and more to a digital terrestrial broadcast network. These technologies have generally evolved from legacy architectures with little real attempt to achieve harmonisation or interworking. Whilst today’s terminals incorporate multiple access technologies, the strategies for connecting to IP networks generally result in simple selection of one of the available networks. For example many mobile handsets have both 3G and WiFi interfaces, but the user has to choose between the networks one at a time. The potential for simultaneously using the available access technologies in a multi-link manner has yet to be exploited. Additionally, requiring users to manually switch between different access networks and maybe also to manually specify which applications may use which networks at which time is clearly impractical. In mobile scenarios the characteristics of the networks may also change rapidly, making manual control not only impractical but also impossible. However, users still need a way to control network usage, due to for example cost considerations. This reasoning illustrates a need for automatic adaptation where user preferences are taken into account.
Figure 1. In a multilink scenario a user can connect to the network using several links simultaneously
Now, is it beneficial to use several networks simultaneously? Yes it is and MARCH will show why both the user and the network service providers benefit and how it can be achieved in the practical world. For example the user can be offered a significant increase in capacity and the broadband network can reach further out. The project aims at convergence of broadband, mobile, and broadcast approaches and its application to urban and rural deployment. In short MARCH addresses key issues for ubiquitous broadband access with a particular focus on multimedia service provision and network architecture effective for connecting all types of users, either fixed or on the move.
1.2 State-of-the-art situation
Combining several access network links for enhancing performance (back-up, throughput, etc) is not new. Channel bonding (as exists in Ethernet switches), layer 2 (link) aggregation and even layer 3 (IP) aggregation have been considered but with a lot of drawbacks. The former method is attached to one technology only; the second is for one Internet Service Provider only and the latter is a partial solution based on IP multi-homing without considering inbound traffic.
The use of multilink in the Internet is known but is generally limited to selecting a single path, with the alternative paths being used in the case of transmission failure. Three approaches are however of particular relevance:
· Recent work in the IETF Monami6 group has extended the Mobile IP protocol to encompass multiple interfaces on the terminal. This is achieved through establishing several tunnels between the terminal and a network agent. Such an approach could enable the terminal to distribute multiple communications over several tunnels. However, the notion of using several tunnels for one service is not envisaged.
· Multi-homing protocols (SHIM6, HIP) allow each end of a communication link to know the pool of addresses of the other end of the link. The simultaneous use of these addresses is however not envisaged, the approach targeting address switching in the event of a link failure.
· SCTP proposes a transport layer approach whereby all possible addresses are exchanged at the start of a connection. Once again the using multiple links for a given communication is not envisaged.
The always best connected concept as envisaged in [1], [2], [3] consists of selecting one or more access networks among a plurality of access networks for connecting the multi-mode terminal with any other terminal or service supporting or not the always best connected architecture. Connecting heterogeneous networks and devices at various locations will enhance the user experience in terms of improved service performance, perceived Quality of Service (QoS) and reduced costs
Both ITU [4] and ETSI [5] have conducted standardisation work aiming at modular scalable network within their next generation network (NGN) initiative. There are also other initiatives for modularity, such as open base station architecture initiative (OBSAI) [6] and similarly for Worldwide Interoperability for microwave access (WiMAX) [7]. These inactivates are ongoing work, but they are not covering the multilink issues so far. Also monolink network solutions are today lacking modularity.
Previous research has addressed several issues like the overall architecture, the end to end QoS, signalling and advertising and the link selection based on user profile, economical inputs (e.g. price) and network status. [2] focuses mainly on an end to end QoS framework providing a consistent QOS whatever the number of [wireless] hops crossed by the end-to-end bearer. [3] focuses on a middleware architecture that abstracts the usage of different modules for adapting the communication to the network interface.
A number of European and North American projects have also envisaged such multi access network architecture:
Ambient Networks [8] introduces a multi-access functionality by specifying a multi-radio resource management (MRRM) entity that can manage the handover acting as the network interface selector. It is present in the terminal and also in the network and controls the forwarding point (FP) that realizes the mapping between high level flows (e.g. HTTP/TCP) to access network flows (e.g. UMTS/3G ). Such mapping results in an active access flow that operates an access network path temporarily until the next handover occurs, ordered by the MMRM. In a particular architecture embodiment, the terminal can also operate the multi radio transmission diversity (MRTD) at different levels including IP level. The transmission diversity consists basically into sending/repeating a data packet several times over several access interfaces. However [8] also envisages the case where packets are sent using, in parallel several interfaces for increasing the bandwidth. MARCH will address this issue with different basic requirements and assumptions including network architecture composed of different access network operators, supporting and extending mobile IP, focusing on video applications, etc.
Multinet [9] aims at designing an always best connected network infrastructure. As with Ambient, Multinet focused on a QOS framework and the network interface selector. They have based their work on NEMO [10], a multi-homing router that acts as a front-end for handling multi-interface management allowing the terminals to operate the infrastructure seamlessly. Although the router can be considered embedded into a terminal this solution is suitable for mobile networks like in trains, automotive contexts.
Daidalos [11] deals with heterogeneous networks and aims at unifying the access to networks whatever their characteristics (including one way broadcast network). Besides building their architecture on IPv6, they have taken various directions ranging from mobility management, virtual identity to broadcast/unicast inter-working. However the typical scenario of using concurrent access networks is not part of the scope.
The Tetherless Computing [12] concept enables communications to be maintained under varying access network conditions through the implementation of a proxy enabling multi-links to be established in an opportunistic manner.