互补随机子梯度分布式认知无线电功率分配

Complementary Stochastic Subgradient Distributed Power Control in Cognitive Radio

频谱作为一种有限资源,随着无线通信服务和器件的日益增多,频段越来越稀缺,认知无线电是有效提高频谱利用率的解决方法之一.认知用户利用授权系统的固有反馈信息,合理设置认知用户的发射功率,两个通信系统能够同时工作,提高了频谱效率,并控制对授权用户造成的干扰.在不需要主动合作的前提下,该文提出了适合认知无线网络的互补随机子梯度分布式功率控制算法(CSDPC),分析了该分布算法的收敛特性,采用互补式的搜索方式,获得了提高收敛速度的有效途径.仿真结果表明,该算法和相关文献比较,收敛时间不到类似算法运算时间的1/10,增强了系统的灵活性,提高了通信容量.

Frequency spectrum is a limited resource for wireless communications and becomes scarce. As a potential solution to use the wireless spectrum resource more efficiently, cognitive radio may control transmitter power to achieve efficient usage on the spectrum resource with protection constraint on the PU QoS based on inherent feedback information in PU communication. In this paper, complementary stochastic subgradient distributed power control algorithm is proposed for spectrum sharing CR system. We investigate the conver gence of the distributed power control algorithm, and find that the algorithm can converge more quickly. The simulations validate the effectiveness of the proposed algorithm by comparing it with similar power control paradigm.