Study On Spectrum Allocation Algorithm Based On Game Theory

Idle spectrum sharing in cognitive radio system is mainly concerned to the leasing price game among main user systems. Game theory, as a important mean to analyze the decision problem, put forward a lot of new spectrum sharing algorithms based on game theory in recent years. But these algorithms have limitations:some regard idle spectrum demand of cognitive users as fixed value without neither spectrum difference analysis nor considering the main users spectrum utilization or the influence of spectrum utilization efficiency of cognitive users. Based on the above problems, this paper improved the existing spectrum allocation algorithm.First of all, this paper summarized the research status and significance of spectrum sharing and game theory in cognitive radio system, and discussed the feasibility of using game theory to study the spectrum sharing in cognitive radio system.Secondly, based on the game theory, the game between the primary users were studied adopting the Bertrand game model. Through spectrum sharing algorithm, the price game between the main users were simulated. The simulation results show that the game converges to the Nash equilibrium, which proved that the the best strategy to achieve the most primary user effectiveness is the game of both sides to achieving the Nash equilibrium, and it is proved that game theory is effect to solve the problems of spectrum sharing.Next, because of the existence of idle spectrum differences in the level and the vertical aspects in the primary users, this paper put forward the concept of cognitive user preferences for the spectrum to represent the influence that level difference of the primary users has on the user loan spectrum considering that the existing algorithms did not consider the influence of level difference of primary users'idle spectrum. The main users offer different prices according to the their preferences. Taking the main user spectrum utilization rate and rate effects on cognitive user utility into account, this paper presented an improved utility function and primary users'charging function, which combined the improved primary users' charging function and cognitive users effect function as a non-cooperative spectrum allocation algorithm. The algorithm can better reflect the fairness, thus the idle primary users spectrum which cognitive users win suit their own command. Theoretical analysis and simulation results verify the effectivenessof the improved non-cooperative spectrum allocation algorithm.Finally, a joint algorithm was proposed. The cognitive users'utility function and primary charging function were from Non cooperative game algorithm, while the corporative game process were from cooperative game algorithm. Simulation results show that joint allocation algothrim in this paper is better than non-corporative. The benefits of cognitive users increase further.