低数据量下的雷达信号脉内调制识别方法  

Method of intra-pulse modulation identification of radar signal with insufficient data samples

作　　者：富炜皓 苟晓鸣 武者东[1] FU Weihao;GOU Xiaoming;WU Zhedong(China Research Institute of Radiowave Propagation,Qingdao 266107,China)

机构地区：[1]中国电波传播研究所,山东青岛266107

出　　处：《现代电子技术》2023年第3期57-61,共5页Modern Electronics Technique

摘　　要：针对部分脉冲宽度窄、采样点数少的雷达信号监测应用场景,提出一种新的雷达信号脉内调制自动识别算法。采用小波变换与瞬时自相关相结合的方法,提取雷达信号瞬时频率特征,根据不同调制方式雷达信号的特征差别,实现了对常规信号、线性调频信号、非线性调频信号、二相编码信号、四相编码信号、二进制频率编码信号和四进制频率编码信号七种不同调制方式雷达信号识别。在不同数据量条件下,仿真结果表明,识别准确率随着码元个数或单个码元采样点数的减少而降低,此识别算法对信噪比为20 dB、采样点数为100个的雷达信号,可以实现对不同调制方式的准确识别,识别准确率高于80%。与传统识别方法相比,该方案所需数据量较少,识别准确率高,易于工程实现。A new method of intra-pulse modulation automatic identification of radar signal is proposed for some radar signal monitoring devices with narrow pulse width and small number of sampling points.The instantaneous frequency features of radar signal are extracted with the method that combines with the wavelet transform method and instantaneous autocorrelation method.According to the feature difference of radar signal in different modulation modes,the recognition of seven radar signals in different modulation modes(namely normal signal,LFM signal,NLFM signal,BPSK signal,QPSK signal,BFSK signal and QFSK signal)has been realized.The simulation results got under the condition of insufficient data samples show that the identification accuracy decreases as the number of symbols or the number of sampling points are decreased,and the recognition algorithm can realize accurate identification of different modulation mode radar signals whose SNR is 20 dB and sampling points are 100.Its identification accuracy rate is higher than 80%.In comparison with traditional identification methods,this method requires less data,has high identification accuracy,and is easy to implement in engineering.

关 键 词：雷达信号识别 脉内调制 自动识别 时频分析 特征提取 准确率统计 

分 类 号：TN911.7-34[电子电信—通信与信息系统]