几种雷达辐射源信号降噪算法比较分析  

Comparison and Analysis of Several Denoise Algorithms for Radar Emitter Signals

作　　者：吴惟诚 潘继飞[1] 刘鑫[1] 

机构地区：[1]电子工程学院

出　　处：《舰船电子对抗》2017年第3期39-44,72,共7页Shipboard Electronic Countermeasure

摘　　要：基于波形配对的雷达辐射源信号分选识别方法受噪声影响大,需要对信号进行降噪处理,降噪效果直接影响分选的准确率和实时性。首先介绍了基于快速傅里叶变换(FFT)的滤波降噪法,对4种经典滤波器进行特性分析。鉴于经典滤波法存在适用范围窄、降噪效果不理想的缺陷,又介绍了模极大值法、阈值法、小波包分析法3类基于小波变换的降噪算法,提出了以纯净信号和降噪后信号的信噪比、均方误差指标来衡量降噪效果,为小波变换降噪法的研究指明了方向。The classification and recognition methods of radar emitter signals based on waveform matching is greatly affected by the noise,so denoise processing to signals is needed,and the denoise effect directly affects the accuracy and real-time performance of the sorting. This paper firstly in-troduces the filtering denoise method based on fast Fourier transfom （FFT） ,and analyzes the char-acteristics of four classical filters. In view of the disadvantages of classical filtering method ： applica- tion scope is narrow and denoise effect is not ideal, this paper introduces three denoise algorithms based on wavelet transform： modulus maxima method, threshold method, wavelet packet analysis method, proposes to use signal to noise ratio and mean square error criterion of pure signal ＆ de- noised signal to measure the denoise effect, which points out the direction for studying wavelet transform denoise method.

关 键 词：降噪 经典滤波 小波变换 阈值 

分 类 号：TN971.1[电子电信—信号与信息处理]