基于EEMD降噪和模糊函数奇异值向量的雷达辐射源信号识别算法  

An Identification Algorithmfor Radar Emitter Signals Based on EEMD Denoise and Ambiguity Function Singular Value Vectors

作　　者：吴力华 杨露菁[1] 袁园 WU Li-hua;YANG Lu-jing;YUAN Yuan(College of Electronic Engineering,Naval University of Engineering,Wuhan 430033,China;Luoyang Electronic Equipment Test Center,Luoyang 471000,China)

机构地区：[1]海军工程大学电子工程学院,武汉430033 [2]中国洛阳电子装备试验中心,河南洛阳471000

出　　处：《火力与指挥控制》2022年第2期121-126,共6页Fire Control & Command Control

摘　　要：针对目前基于模糊函数提取的几何学特征,在低信噪比时表征能力弱,导致识别准确率不高的问题,提出了一种基于EEMD降噪和模糊函数奇异值向量的识别方法。选取合适的EEMD参数,对时域信号进行降噪,提取模糊函数矩阵奇异值向量,求解其交叉熵作为特征,实现雷达辐射源信号识别。仿真实验表明,信噪比大于-5 dB时,所提方法对于BPSK、QPSK、MSEQ、BFSK、FMCW和LFM-BC六类典型调制信号,能达到大于90%的平均识别准确率,基本满足实际复杂电磁环境需求。The geometric features abstracted based on ambiguity functions show the weak representation ability under low signal noise ratio(SNR) leading to the problem of low identification accuracy.An identification algorithm based on EEMD denoise and ambiguity function singular value vectors is proposed.First,the appropriate EEMD parameters are chosen to denoise the time domain signals.Then,the ambiguity function matrix singular value vector of each signal is extracted.Finally,the solution of cross entropy is used as feature to achieve the identification of radar emitter signals.The simulation experiments show that the average identification accuracy rate of six kinds of typical modulated signals,i.e.,BPSK,QPSK,MSEQ,BFSK,FMCW and LFM-BC,by proposed algorithm reaches above 90%when the SNR is above-5dB and it can basically meet the actual complex electromagnetic environment requirements.

关 键 词：信号识别 EEMD 模糊函数 奇异值向量 交叉熵 

分 类 号：TN974[电子电信—信号与信息处理]