一种适应低信噪比的脉内识别算法  

An intrapulse identification algorithm at low SNR

作　　者：曹俊纺 张玉喜 吴昊 臧勤 CAO Jun-fang;ZHANG Yu-xi;WU Hao;ZANG Qin(No.8 Research Academy of CSSC,Nanjing 211153)

机构地区：[1]中国船舶集团有限公司第八研究院,南京211153

出　　处：《雷达与对抗》2022年第2期7-10,43,共5页Radar & ECM

摘　　要：针对传统的雷达信号脉内调制类型在低信噪比下识别能力不高的局限性,提出一种改进的适应低信噪比的信号调制类型分类识别算法。该算法对雷达脉冲信号流进行时间和脉宽格子划分,对同一格子中脉冲信号采用一阶差分自相关函数法计算特征向量,根据特征向量值进行聚类统计分类。分类后对每组内脉冲信号进行平方处理,再次计算特征向量并进行聚类统计分类,选择每组内质量较好的脉冲信号进行调制类型识别,提高了信号分类识别正确率,通过仿真实验验证了该方法的有效性。In view of the limitation that the conventional radar signal intrapulse modulation types cannot be accurately identified at low SNR, an improved classification and identification algorithm is proposed for signal modulation types at low SNR. The algorithm divides the radar pulse signal flow into time and pulse width grids, uses the first-order differential autocorrelation function method to calculate the eigenvectors of the pulse signals in the same grid, and performs clustering and statistical classification according to the eigenvector values. After classification, the pulse signals in each group are squared, the feature vectors are calculated again, clustering and statistical classification are performed, and the pulse signals with better quality in each group are selected for modulation type identification, improving the accuracy of signal classification and identification. It is verified that the method is effective through the simulation test.

关 键 词：脉内调制类型识别 低SNR 识别准确率 

分 类 号：TN957.51[电子电信—信号与信息处理]