协作频谱感知系统中的结盟技术研究

Dynamic Coalition Formation in Cooperative Spectrum Sensing System

针对基于联盟的协作频谱感知的技术,侧重研究如何结盟使得感知性能最佳。首先建立了基于主用户的电磁传播环境模型,并根据博弈论的思想,构造了结盟性能评价价值函数,以此来判断各种不同的结盟方式的性能优劣。然后针对无线通信环境下存在的不同噪声和用户不同的地理分布,设计了多种联盟模型,进行了不同结盟场景构思和算法推导。大量的仿真实验表明,采用联盟协作频谱感知,不仅降低了传统的多用户协作的复杂度,而且其感知性能比集中式融合好。最后根据大量实验总结了多用户联盟频谱感知的结盟规律和最优组合。

For cooperative spectrum sensing technology,we focus on how to form the dynamic coalition structures to gain the best performance.First,we build a primary user based on electromagnetic propagation model.Then,we construct the value function according to the game theory,to judge the performance of different coalition structures.Finally,we design various coalition structures according to various noise and various user distribution under wireless communication circumstances.Experimental results show that the proposed method not only makes traditional multi-user cooperation easier,but also has a better perceived performance.Based on the experiments,we give the coalition rules and an optimized coalition structures for the multi-user spectrum sensing.