基于MLE的频谱感知统计量算法

摘要

提出一种基于认知用户接收与主用户之间的信道特性为参量未知的高斯分布,构造主用户的信息特征参量完全未知下的频谱感知统计量,该统计量主要基于广义似然比检验(GLRT).对信号的协方差矩阵进行极大似然估计.理论分析和仿真结果表明该统计量计算简单,具有很强的实用性.%For the cases that the information of the primary users is completely unknown to the secondary users, and the channel characteristics between the primary users and secondary user are also unknown, a kind of statistics is constructed for spectrum sensing. The statistics are constructed based on the generalized likelihood ratio test (GLRT), and with the developed algorithm the cognitive user constructs a covariance matrix based on the eigenvalue statistics through the signal covariance matrix of the maximum likelihood estimation. The theoretical analysis and the simulation results indicate that the statistical expression is simple and practicable.