基于模仿机制的认知无线电频谱动态接入

In cognitive radio networks (CRNs), primary users (PUs) and secondary users (SUs) adopt either cooperative or non-cooperative behaivors in order to improve network performance. In the system model we consider, each SU selects whether to cooperate with the PU or not, which is formulated as two strategies. In order to study the optimal choice of each user, we propose an imitation mechanism and we further prove the converging property of the system if applying this mechanism. Finally, we propose two algorithms: One centralized algorithm shows the converging property of the dynamic replication function. And compared with the centralized one, the other distributed algorithm which enables each user to activate imitation process provides better performance without requiring broadcast information from the central controller. Simulation results prove that the imitation mechanism efficiently makes the system converge to an equilibrium point during the process.