基于小波——神经网络的雷达信号脉内特征分析  

Extraction of the In-pulse Characteristics of Signal Based on Wavelet-Neural Network Measurement

作　　者：孔辉 王杰贵[2] 

机构地区：[1]61251部队,河北秦皇岛066102 [2]电子工程学院,合肥230037

出　　处：《电子对抗》2008年第5期15-21,共7页Electronic Warfare

摘　　要：时频分析是脉内特征分析中常用的一种方法，通过提取雷达脉冲信号的小波脊函数可以获得脉内调制信号在每个采样点的瞬时频率，然后再使用神经网络的方法对小波变换提取出的瞬时频率进行分析，便可以得到信号的脉内调制特征。以MoBet小波和多层感知器为例，对低信噪比下的典型雷达信号进行了仿真分析，结果验证了小波-神经网络分析脉内特征的有效性。Time-frequency is a common measurement in in-pulse characteristics analysis. Instantaneous frequency can be acquired by distill wavelet-ridge in radar pulse signal. Neural Network is used to analyze the instantaneous frequency to extract the in-pulse characteristic. An improved method which analyzes radar signal through morlet-wavelet and muhilayer perceptron is presented. Simulations are performed under the strong noise circumstance. Result proved that the extraction of the in-pulse characteristics by combined frequency measurement is effective.

关 键 词：小波分析 神经网络 脉内调制 特征分析 

分 类 号：TN957.51[电子电信—信号与信息处理] TP183[电子电信—信息与通信工程]