宽带压缩频谱感知非稀疏保护策略

Non-sparsity protection strategy for compressed spectrum in wideband cognitive radio

针对频谱非稀疏情况下宽带压缩频谱感知性能下降的问题,提出一种基于信号重构模型参数分析的频谱非稀疏保护策略,该策略根据次用户接收到的相邻三帧信号之间的相关系数的大小动态判断宽带压缩频谱感知正确与否.首先建立基于高斯分层概率的宽带压缩信号模型对压缩重构进行分析,在此基础上综合考虑重构各帧信号模型参数的相关系数与频谱稀疏性之间的关系,设计一种认知用户能够根据信号模型参数相关系数的大小制定频谱稀疏与否的判决条件,进而制定保护措施.仿真结果表明:所提保护策略可以有效检测到由频谱非稀疏导致的宽带压缩频谱感知错误,降低对主用户的干扰概率.

In order to overcome the failure of compressed spectrum sensing （CSS）in wideband cogni-tive radio owing to the non-sparsity of the spectrum,a protection strategy based on the channel pa-rameters analysis of the spectrum was proposed.The second user could j udge the compressed spec-trum sensing failure according to correlation coefficient of the received adj acent signals.A Gaussian process framework was set to model the compressed spectrum reconstruction and the relationship be-tween the correlation coefficient of the model parameters and the sparsity was analysed.The results of simulation indicate that the proposed non-sparsity protection strategy of compressed spectrum sensing （NPCSS）can effectively detect the failure of CSS and the interference probability to primary users has been declined.