基于矩阵二值化的频率捷变雷达信号分选  

Frequency Agility Radar Signal Sorting Based on Matrix Binarization

作　　者：郭利荣[1] 何明浩[1] 郁春来[1] 王冰切[1] 

机构地区：[1]空军预警学院,湖北武汉430019

出　　处：《雷达科学与技术》2014年第6期574-579,共6页Radar Science and Technology

基　　金：国家自然科学基金(No.61201123);空军装备科研项目(No.KJ2012232)

摘　　要：针对频率捷变雷达信号的分选,提出了一种基于时频矩阵二值化的频率捷变雷达信号分选新方法。该方法首先对信号进行时频变换,得到时频矩阵;然后对时频矩阵二值化处理,提取信号在时频域能量分布归一化值作为信号的相参特征;最后采用支持向量机分类器实现信号分选。文中对频率捷变雷达信号进行仿真实验,结果表明,该方法在较低信噪比下仍能获得较为满意的分选准确率,当信噪比为6dB时,信号分选准确率达到96.17%,验证了所提出方法的有效性。To sort frequency agility radar signal correctly,a novel frequency agility radar signal sorting approach based on the time-frequency distribution matrix binarization feature is proposed.Firstly,the time-frequency transform for radar signals is made to get the time-frequency distribution matrix.Then the binariza-tion processing for time-frequency distribution matrix and the signals energy distribution normalized values in time-frequency domain are extracted as the coherent characteristic.Finally,the support vector machine is applied to signal sorting.Simulation results show that the proposed approach can achieve satisfactory accurate sorting in low signal-to-noise rate（SNR）.Even for SNR=6 dB,the overall correct sorting rate of advanced frequency agility radar signals is 96.17%.The validity of the approach is demonstrated by experimental results.

关 键 词：时频分析 矩阵二值化 相参特征 支持向量机 信号分选 

分 类 号：TN971[电子电信—信号与信息处理]