认知无线电网络中的分布式协作感知算法

摘要

针对认知无线电中频谱感知的问题,讨论了分布式协作感知场景下最优的感知算法.介绍了经典的基于放大前传(AF)、解码前传(DF)和选择中继(SR)的分布式协作方案,包括模型、检测概率和虚警概率.提出了一种新的分布式协作感知算法,该算法将原始统计量和信噪比发送给协作用户,减少信道噪声对最后决策结果的影响.运用似然比准则推导出了最优的加权系数,以及决策函数.对几种协作方案的中断概率进行理论推导,并用数值仿真的方法验证了算法在中断概率上的性能.仿真结果表明,提出的方法在检测概率和中断概率上远优于其他分布式协作方案,但该方法需要协作用户间无损耗的传输,这也给实际的应用提出了很大的挑战.%This paper discussed optimal distributed cooperative spectrum sensing in cognitive radio networks. It first introduced the conventional AF,DF and SR protocols including the system model,detection probability and probability of false alarm. And proposed a novel optimal distributed cooperative spectrum sensing ( ODCSS) algorithm which was based on the observation energy and the signal to noise ratio (SNR). It analyzed all the protocols. The extensive simulations show that the ODCSS algorithm outperforms the other protocols and can be utilized by the practical system.