Research On Wireless Resource Allocation Algorithm Based On Game Theory

Cooperative relay, as a promising technique of wireless communication network, can both drastically improve transmission performance and expand the coverage area. But for real networks, the component nodes are generally selfish. These nodes do not contribute their resource to relaying. In this kind of selfish cooperative systems, the performance of cooperative communication depends largely on resource optimum allocation. Thus we must establish sound incentive mechanism to make selfish nodes initiatively participate in cooperation. Game theory studies how selfish individuals compete and cooperate with each other, and this has been proved an effective way to address resource allocation issue in selfish wireless cooperative relay networks. In this thesis, for power allocation in three asymmetric fundamental network models of cooperative relay, we study the algorithms relevant based on game theory and the agreement of amplify and forward, giving specific resource allocation strategy.System with multi-source and single-relay:We consider not only the selfishness of relay, but also that between source nodes, improving the commonly used resource pricing strategy in current research and giving different strategies for multi-source. The problem of allocating relay power resource for relay can be formulated as a non-cooperative game between source nodes, and the Nash equilibrium solution can be solved by a distributed algorithm. The answer is the best allocation strategy without considering Pareto efficiency, the improved Pareto dominant solution maximizes the utility of all source nodes. Finally, simulations prove the proposed solution could effectively achieve fairness in multi-source and optimize system performance.System with single-source and multi-relay:Relay nodes compete as sellers, so the single-source has priority to set resource price for relay. For allocating relay power resource between relay nodes, we can formulate it as a Nash bargaining problem of cooperative game between them. The previous researches basically are on the background of understanding of mutual information between different nodes, this situation is ideal. We aim at incomplete information and study the asymmetric Nash bargaining allocation strategy. The solving of Nash bargaining solution ultimately transforms into solving Lagrange multiplier that we can rely on gradient descent algorithm to work out. Simulations show that resource allocation could achieve a good tradeoff between efficiency and fairness in different information situations.System with multi-source and multi-relay:A kind of relatively complicated system which can be studied on the basis of the first two systems. Considering the payment and reward between source and relay nodes, which allows us to refer to buying and selling game models in economics and take Walrasian equilibrium strategy. Depending on proposed ascending-clock auction algorithm, the Walrasian equilibrium price of relay resource can be solved. Simulation results prove the existence of equilibrium, maximum use of relay resource and the optimal system performance.