基于EEF准则的认知无线电合作频谱感知

Cooperative spectrum sensing for cognitive radio based on exponentially embedded family criterion

频谱感知是认知无线电的一个重要组成部分。提出了一种基于指数嵌入族(exponentially embedded family,EEF)准则的合作频谱感知算法。与传统方式下所有合作用户皆参加检测的方法不同,提出的算法依靠用户选择机制,并且不需要知晓授权用户信号的任何先验信息。该算法先对参与合作频谱感知的认知用户进行筛选,然后在筛选出的最优用户频谱观测数据的基础上生成全局检测统计量(global decision statistic,GTS),最后对授权用户是否存在做出全局判决。仿真表明,在虚警概率保持不变的情况下,进行最优用户选择(optimal user selection,OUS)的合作频谱感知算法的检测概率优于未进行最优用户选择的算法。

Spectrum Sensing is a fundamental component in cognitive radio. We propose a cooperative spectrum sensing algorithm via exponentially embedded family criterion. Compared with the traditional spectrum sensing algorithms,the proposed algorithm depends on user selection mechanism and does not require all the cognitive users to participate in data fusion. In addition,it does not need a priori knowledge of the primary user signal. The proposed algorithm selects the optimal users firstly,and then based on the spectrum observations of the selected users,the fusion center generates the global decision statistic（ GTS） to determine whether the PU signal exists or not. The simulations demonstrate that under the criterion of constant false alarm probability,the proposed optimal user selection（ OUS） algorithm yields improved detection probability compared to those without OUS.