|
|
Multi-Element Multihop Backhaul Reconfigurable Antenna NEtwork
|
The Multi-Element
Multihop Backhaul Reconfigurable Antenna Network
(MEMBRANE) project aims to bring an efficient wireless backhaul
design as an alternative technology to serve wireless broadband
networks in cases where a wired backhaul would be more costly
to access and/or would take longer to deploy.
|
At
a glance: MEMBRANE
Project Coordinator
Professor Kin Leung
Imperial College London
Tel: +44 20 7594 6238
Fax: +44 20 7594 6328
Email kin.leung@imperial.ac.uk
www.imperial.ac.uk/membrane
Partners: Imperial College (UK), Lucent (UK), ETH Zurich (CH),
Intel (RU), CEFRIEL (IT), Intracom (GR), Telefonica (ES)
Duration: January 2006-June 2008
Total Cost: 4.3 M EUR
EC Contribution: 2.8 M EUR
|
Main Objectives:
The ongoing proliferation of wireless broadband data services
is expected to lead to increased needs on the side of the
backhaul network, which transports data between the access
network and the wired Internet, as well. The typical upgrade
of wired lines to high-speed fibre networks is not always
an available or economically attractive solution. In such
cases, wireless alternatives could offer an appealing alternative.
We propose the design of efficient wireless backhaul networks
that meet the Quality of Service (QoS) demands of high speed
access wireless networks, thus providing a technology shortcut
that will help satisfy the social need for broadband data
anytime anywhere in a much more expeditious way.
Given the need for an efficient wireless backhaul network,
its successful deployment necessitates careful design and
this is likely to require nothing short of a number of technological
breakthroughs. The main envisioned requirements of such backhaul
network are the following:
|
Quality of Service:
Throughput performance: in order to satisfy the growing demands
of data throughput of wireless access networks, a primary
target of the corresponding backhaul network is its high capacity.
Delay performance: several applications have stringent delay
requirements whose satisfaction is critical to acceptable
user experience.
Coverage: especially in the case of remote or isolated areas,
it is important that the backhaul network has enough range
to reach the end nodes.
Overall capacity: obviously, while providing performance and
coverage, being a wireless network, such a backhaul has to
be able to provide sufficient capacity from a system point
of view.
Reconfigurability: another important feature
of wireless networks is that they must be capable of reconfiguring
themselves, in terms of network topology, traffic flow and
propagation conditions.
Heterogeneity: since wireless broadband access
networks employ several technologies, it is important that
the proposed backhaul network is capable of matching these
heterogeneous networks.
Ubiquity: The addition of efficient wireless
backhaul will go a long way in ensuring the delivery of wireless
broadband access to the entire addressable population in a
ubiquitous manner.
Openness: from the operator's point of view,
it is important that the developed concepts meet the requirements
of double-openness, i.e., standard interfaces and multi-supplier
technologies.
|
Enabling Technologies
To satisfy the above requirements, a number of enabling technologies
will be exploited that have the potential, when combined,
to deliver the expected performance:
Multi-hopping, i.e. the use of relays between
the end user and the end node, is primarily motivated by the
low power and the low heights of the access (AN) and relay
nodes. Clearly, in low power transmissions, multi-hopping
helps increase the range. Moreover, since low height ANs are
likely to be surrounded by several obstacles, multi-hopping
helps avoid the problem via multiple links that are more likely
to have LOS between them.
Intelligent Antennas (IA) such as beamforming
(BF) and MIMO transmission can offer improved throughput and
range. Moreover, through their added "spatial"
degrees of freedom, they can also reduce interference from
adjacent links. Reconfigurable IA algorithms can be designed
that match in the best possible way a given propagation environment,
boost throughput and reduce interference thus improving overall
end-to-end performance. The combination of multihop with multiple
antennas is a field that is virtually unexplored and it will
constitute a major innovation in MEMBRANE.
Opportunistic routing & scheduling targets
primarily the satisfaction of the network's delay requirements;
when combined with IA it can also increase the end-to-end
throughput. By opportunistic routing and scheduling we refer
to routing and scheduling policies that take into account
the network context. By this we mean the quality of different
links, the data traffic needs on different ANs, the capacity
of different relays, the packet activity of different links,
etc. Another potential gain of multi-antenna multihop mesh
networks is their potential for significantly improved QoS
via combined opportunistic routing and spatial data multiplexing.
|
REFERENCE ARCHITECTURE
Technical Approach
The MEMBRANE project is structured in the following workpackages:
WP1: Project Management
WP2: Problem definition and scenarios
WP3: Theoretical multihop network performance analysis
WP4.1: Multi-antenna link signalling and routing
WP4.2 Routing, scheduling and power control for wireless
backhaul network optimisation
WP4.3 Wireless multihop backhaul IP network design
WP5.1 MEMBRANE System Simulation Platform
WP5.2 MEMBRANE Prototype
WP6.1: Overall performance evaluation
WP6.2: Exploitation and Dissemination
Anticipated outcome and impact
The studies within MEMBRANE will:
Enhance the understanding of the fundamental capacity limits
of multiple antenna multihop networks.
Provide optimal network topologies, in demanding propagation
conditions and under low power or low cost constraints.
Develop efficient, reconfigurable routing and scheduling
algorithms that maximize QoS, power utilization and service
fairness. The joint routing and scheduling algorithms will
represent a new approach to routing and scheduling, which
fully exploits the knowledge of condition of radio links
along any given route for the best route selection and scheduling
performance in multihop wireless networks.
Develop reconfigurable smart antenna techniques for throughput
maximization and delay minimization.
Invent new techniques for efficient resource utilization
by combining intelligent antenna, routing and scheduling
techniques.
Provide proofs-of-concept of the developed network design
and algorithms via a comprehensive system-level evaluation
platform and proof-of-concept demonstration.
Give recommendations for next generation air interface design
for the backhaul network, that are based on cross-layer
optimized design and its novel spatial/temporal/opportunistic
routing and scheduling policies.
Download this Project Summary as a PDF 
|
|
|