摘要
针对在低信噪比特别是参有混合调制信号环境下雷达辐射源信号识别率低的问题,提出了一种基于时频能量分布的辐射源信号特征提取和识别方法.该方法首先对信号进行时频变换得到时频分布,经过一系列图像预处理后得到时频能量分布;然后通过行相加变换取平均提取信号特征向量,利用支持向量机分类器实现信号分类与识别;最后对6种典型雷达信号进行了仿真.仿真结果表明,本文方法在低信噪比条件下且有混合调制信号时识别效果较好,在信噪比为-6 dB时平均识别率仍能达到84%.
Aiming at the problem of low recognition rate of radar emitter signals under the condition of low signal-to-noise ratio(SNR),especially with mixed modulation signals involved,this paper proposes a method of feature extraction and recognition of emitter signal based on time-frequency energy distribution.Firstly,the time-frequency distribution is obtained by time-frequency transformation of the signal,and then the time-frequency energy distribution is obtained after a series of image preprocessing.Secondly,the signal feature vector is extracted by the average of the row addition transformation,and support vector machine(SVM)classifier is used to realize signal classification and recognition.Finally,six typical radar signals are simulated.The simulation results show that under the condition of low SNR and mixed modulation signals,the recognition effect is better,and the average recognition rate can still reach 84%when the SNR is-6 dB.
作者
伍建昌
屈翼展
陈新
胡乔林
WU Jianchang;QU Yizhan;CHEN Xin;HU Qiaolin(Air Force Early Warning Academy,Wuhan 430019,China)
出处
《空军预警学院学报》
2020年第1期31-34,共4页
Journal of Air Force Early Warning Academy
关键词
时频能量分布
时频变换
支持向量机
雷达辐射源识别
time-frequency energy distribution
time-frequency transformation
support vector machine(SVM)
radar emitter recognition
