An Evolutionary Game-theoretic Approach to Spectrum Sharing System

认知无线电是一种能够学习周围环境并自适应传输的无线通信技术，因其良好的性能而备受关注。认知无线电的一项使能技术是频谱共享。通过引入进化博弈理论对频谱共享过程中多个决策主体的动态行为进行了建模。首先，利用进化博弈理论对频谱共享系统进行了建模，阐明了频谱共享系统的特点。其次，针对频谱共享系统中的干扰问题，提出了基于少数人博弈的分布式传输调度方法，分析表明，由于多个发射机的传输定时是自组织的，因此可以很好地管理干扰。第三，对移动路由器的自适应控制进行了建模，阐明了其收敛条件。

Cognitive radio, which learns surrounding radio environment and adaptively transmits, has attracted much attention due to its promising capability. One of enabling technologies for cognitive radio is spectrum sharing. We focused on the fact that dynamics of multiple decision-making entities in spectrum sharing can be modeled by introducing evolutionary game theory. First, the spectrum sharing system is modeled by using evolutionary game theory and its characteristic is clarified. Second, decentralized transmission scheduling to manage interference in the spectrum sharing system is proposed based on minority game, and it is revealed that the interference is well managed because the transmission timings of many transmitters are self-organized. Third, adaptive control for mobile routers is modeled and its convergence condition is clarified.