Cooperation and Conflict: Coalitional Games in Spectrum Sharing

Project Details

Description

Mandayam, Narayan

Rutgers Univ New Brunswick

Cooperation and Conflict: Coalitional Games in Spectrum Sharing

Cooperative communication techniques hold the promise of promoting efficient spectrum sharing and a good deal of important research has studied approaches such as collaborative signal processing, cooperative coding, relaying and forwarding. What is perhaps less generally considered in research studies is that cooperation may involve significant costs, and that the greatest immediate benefits will not necessarily go to the users that bear the greatest immediate cost. For example, cooperation may require that a terminal delay its own transmissions and use its limited energy budget to relay messages for other terminals. Moreover, even if there were no added cost in terms of energy or delay, the rules by which the benefit is distributed may create yet another type of disincentive to cooperation. Thus, even in the absence of any transaction costs for cooperation, the aggregate benefits of user cooperation can only be realized through techniques that provide adequate incentives to each user, and for each transaction.

This project is using coalitional game theory as a framework to provide significant insight into the dynamics of cooperative communications as well as the incentives through which such cooperation might be achieved. Integrating models from game theory with communications and information theory, the cooperation mechanisms being studied are: (1) coalitions in receiver cooperation, (2) coalitions in transmitter cooperation, and (3) incentives and costs in coalition formation. This project is contributing to the knowledge and technology of coalition formation that can result in improved spectrum sharing in cognitive radio networks. An educational supplement to the theoretical research activities is using the ORBIT Testbed for evaluating simple practical coalition formation mechanisms.

StatusFinished
Effective start/end date3/1/072/28/09

Funding

  • National Science Foundation: $100,000.00

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.