认知无线电中基于无限次重复博弈的功率控制算法  

A Repeated Games of Power Control Algorithm for Cognitive Radios

作　　者：廖鼎[1] 杨震[1] 

机构地区：[1]南京邮电大学信号处理与传输研究院,江苏南京210003

出　　处：《南京邮电大学学报（自然科学版）》2009年第5期72-75,共4页Journal of Nanjing University of Posts and Telecommunications：Natural Science Edition

基　　金：国家自然科学基金(60772062);教育部重点科技项目(206055)资助项目

摘　　要：通过博弈来实现认知无线电中的功率控制。当两用户功率控制博弈时,用户通过功率的迭代注水法实现最大化自身速率,达到纳什均衡。功率分配的纳什均衡点构成囚徒困境,但这种囚徒困境的均衡点并非全局最优。应用两用户功率控制的无限次重复博弈算法,通过选择严厉的触发策略,当折扣因子σ足够接近于1,使两用户在无限次重复博弈中一直合作使囚徒走出了困境,最终达到了功率分配的帕雷托最优的均衡结果。The transmit power control problems in a cognitive radio environment may be viewed as a game theoretic problem. In a two users power control game, each user optimizes its own rate by power iterative waterfilling and reaches Nash equilibrium. The stable equilibrium point of the two users game is suboptimal due to the prisoner＇ s dilemma phenomena. An two users power control infinite repeated games is proposed by choose rigid touch of stratagem, if discount parameter approaches to one. Two users will cooperate in infinite repeated games and avoid prisoner＇ s dilemma. The stable equilibrium point is pareto optimal.

关 键 词：认知无线电 功率控制 博弈论 囚徒困境 帕雷托 

分 类 号：TN925.5[电子电信—通信与信息系统]