基于排队时延及博弈分析的认知无线网络信道分配算法

Cognitive network channel allocation method based on the queuing delay and game analysis

提出了一种适用于认知无线网络的分布式动态频谱资源分配算法。该方法以业务分组的传输成功率作为用户的效用函数,通过优先级排队模型求解传输时延,并采用分布式博弈获得各用户的信道分配策略。与已有的算法相比,所提算法对策略迭代方式进行了改进,且采用了动态策略调整步长。各从用户根据当前感知的网络状态和其余用户的策略,不断动态调整自身的信道选择策略。所提出的算法能够使各认知用户信道选择策略更加快速地收敛到策略均衡点,有效抑制策略的振荡,减小分组丢失率。基于MATLAB对所提出的算法的性能进行了仿真,仿真结果验证了该算法的有效性。

A new dynamic spectrum allocation method for distributed cognitive radio networks was proposed.In the method,packet transmission success rate is chosen as the utility function,the transmission delay is resolved by the priority queuing model,and the user’s channel allocation strategy is achieved by distributed game.Compared with other algorithms,the policy iteration process is improved effectively,and a dynamic strategy adjustment step is adopted.Each network user dynamically adjust their channel choose strategy based on current perception of the remaining user’s strategy.The proposed algorithm lets channel selection strategy converge more quickly to equilibrium point,and effectively suppresses the policy shocks.The simulation results validate the proposed algorithm.