一种基于盲源分离的MIMO雷达侦察识别方法

An MIMO Radar Reconnaissance Identification Method Based on Blind Source Separation

针对集中式MIMO雷达信号分选面临的问题,提出一种新的侦察识别方法,令多侦察机同时在不同方位对MIMO雷达信号进行采样,使其能够获得多个独立的观测量,在此基础上,利用盲分离技术中的复值非高斯性最大化(CMN)算法对MIMO雷达信号进行分选。仿真结果表明,在不考虑幅度起伏的条件下,当SNR达到6 dB时,对相位编码、频分线性调频等典型正交波形的分离指数均低于-10 dB,同时,分离后波形的相关性分析也验证了方法的有效性。

To correctly sort MIMO radar signals, a new tce method is presented, which can make multiple reconnaissance planes collect radar signals in multiple positions at the same time, and multiple independent observations can be obtained. The blind source separation algorithm of complex maximization of non-Gaussianity （CMN） is utilized to sort MIMO radar signals. Simulation results show that a separation performance of - 10 dB can be achieved with SNR as 6 dB for several typical orthogonal waveforms such as phase-coded and frequency division liner frequency modulation waveforms without considering the amplitude fluctuation. The relative analysis of separated waveforms also demonstrates the validity of the method.