Topic: Application of game theory to distributed radio resource management in OFDM based wireless systems.

 

Context

 

Today’s trend in wireless system is to use multi-carrier waveforms which present good efficiency for being used into a radio mobile environment as well as ability to be adaptable dynamically to a particular context, taking into account both local and multi-cell aspects. Such waveforms are sometimes classified as heteromorphic.

Given the waveform flexibility to adapt to user requirements, system constraint and local and temporal characteristic, the way to adapt the waveform or more generally to allocate the resource leads to a complex optimisation problem. More explicit questions could be for instance:

·          Code and power allocation in CDMA systems

·          Carrier allocation for (O)FDMA systems

·          Time slot allocation for TDMA systems

 

Traditionally, such resource allocation is done within a given spectrum. The operator has obtained from the regulator the right to use a given part of the radio bandwidth. Actually, today, the spectrum is allocated to the different radio access technologies in a static fashion. In this respect, some fixed spectrum portions are assigned to different standardised radio access systems that are exclusively operated either by some operators (licensed systems/bands) or by any free users (unlicensed systems/bands).

This regulated approach guarantees that different radio systems will not interfere each-other. However, with the digital convergence, and the emergence of reconfigurable equipments (able to run various waveforms in a wide range of frequencies) such static approach is no longer the most efficient since each radio systems load may vary both spatially and temporally. As a result, one can expect greater efficiency from a system able to allocate resource (e.g. carrier) in the whole available spectrum, as long as each allocation will be done to limit the interference created in the neighbourhood.

Obviously, a centralized entity, having knowledge of all the users, access point, channel conditions etc… could optimize the allocation. This is unfortunately not possible in real systems where the decisions should be taken in a more distributed way. On the other hand a distributed approach could benefit from local information and thus fast adaptation but when each user (or access point) tries to maximize their interest (e.g. the data rate) the resulting system performance are far from being optimal since every actors in the system behave selfishly.

The game theory has been developed to investigate systems where actors interact in a way to maximize their own objective, assuming some knowledge of other actor’s strategy. Therefore, game theory could be applied to the problem of resource allocation in a wireless system in order to derive innovative strategies for distributed RRM, with an efficient signalling.

 

Basically, the purpose of the work is to:

·          Understand the problem of resource allocation in a deregulated context. In a first step, “deregulated” may corresponds to unlicensed bands and the resource are carriers for an OFDMA system.

·          Understand game theory

·          Model the problem to be solved using game theory approach

·          Find/prove the (non)existence of an equilibrium. Map the equivalence of various equilibria concepts (Nash, Wardrop, Stackelberg, etc.) in real wireless contexts

·          Propose simple strategies to regulate the non cooperative game. For instance, introduce pricing component to change the equibrium point.

·          Propose simple strategies for cooperative game

 

Validation of the proposed strategies will be done by analytical or simulation approach. Comparison with “traditional” resource allocation should be done too.

 

Key Words:

 

 

Skills:

 

Duration:

 

Place:

Parc les Algorithmes

Saint-AUBIN

91193 Gif sur Yvette Cedex

 

Contact:

Merouane Debbah

Eurecom Institute - Mobile Communications Dept.

2229 Route des Cretes BP 193

F-06904 Sophia Antipolis Cedex, France

Email:debbah@eurecom.fr            Tel: +33 4 93 00 81 86

http://www.eurecom.fr/~debbah      Fax: +33 4 93 00 82 00

 

References

 

[1] NEEL J., BUEHRER R.M., REED B.H., GILLES R.P., « Game theoretic analysis of a network of cognitive radios », IEEE MWSCAS-2002, p. III-409 . III-412 vol. 3, 4-7 août 2002.

[2] E. Altman, T. Boulogne, R. El-Azouzi, T. Jimenez, and L. Wynterz, “A survey on networking games in telecommunications”, Computers and Operations Research, 2004.