基于特征值分析的通信信号快速盲检测方法

A Fast Blind Detection Algorithm of Communication Signals Based on Eigenvalue Analysis

对中频通信信号的协方差矩阵进行特征值分解,根据特征值在信号子空间和噪声子空间的分布差异,提出了基于特征值分析的快速盲检测方法.推导了信噪比与检测量之间的关系,进而对检测性能进行了理论分析,并分析了算法的计算复杂度.对MPSK、MQAM、MFSK等常用通信信号的仿真实验表明,在虚警概率小于1%、信噪比为-10dB时,盲检测概率均可达到90%以上.

By performing eigen-decomposition（ED）to covariance matrix of intermediate frequency（IF） communication signals,eigenvalues distribution in signal subspace and noise subspace is obtained.According to the distribution difference,a fast blind detection algorithm based on eigenvalue analysis is presented.Relationships between detection quantity and signal-to-noise ratio（SNR） are deduced,and detection performances are analyzed theoretically.Computational complexity is also analyzed.For commonly used communication signals, such as MPSK, MQAM, MFSK, etc, simulation results indicate：when false alarm probability is set to less than 1% ,bind detection probability reaches above 90% at SNR - 10dB.