基于Gabor纹理特征和支持向量机的跳频调制识别  

Modulation Recognition of Frequency Hopping Signal Based on Gabor Texture Feature and Support Vector Machine

作　　者：王冠 孙静 王金宇 张志伟 王晖 WANG Guan;SUN Jing;WANG Jinyu;ZHANG Zhiwei;WANG Hui(A ir Force Communication NCO A cademy,Dalian 116600,China)

机构地区：[1]空军通信士官学校,辽宁大连116600

出　　处：《火力与指挥控制》2023年第9期110-116,122,共8页Fire Control & Command Control

摘　　要：针对跳频信号调制方式识别问题,提出一种基于Gabor变换和支持向量机的调制识别方法,该方法将信号识别问题转化为图像识别问题。利用Choi-Williams分布获得6种跳频信号的时频图像;将Choi-Williams时频图像灰度化处理,应用Gabor变换提取单一尺度4个角度下的纹理特征,通过PCA降维处理形成特征参数;利用多分类支持向量机分类训练、识别。仿真结果表明,所提方法在信噪比-8 dB时,BFSK、BASK、BPSK、QPSK、MSK及16QAM共6种跳频信号调制识别率高达90.67%,相比于其他纹理特征方法,具有更高的识别率及抗噪性。A modulation recognition method based on Gabor transform and support vector machine is proposed for the identification problem of frequency hopping(FH)signal modulation mode,which transforms the signal recognition problem into an image recognition problem.Firstly,Choi-Williams distribution is used to obtain the time-frequency images of six kinds of FH signals;then,Choi-Williams time-frequency images are grayed out and Gabor transform is applied to extract the texture features at four angles of a single scale,and the feature parameters are formed by PCA dimension reduction;finally,the multi-classification support vector machine is used for classification training and recognition.The simulation results show that the proposed method has a high recognition rate of 90.67%at signal-to-noise ratio of-8 dB for a total of six kinds of FH signal modulations of BFSK,BASK,BPSK,QPSK,MSK and 16QAM,which has a higher recognition rate and noise immunity compared with the other texture feature methods.

关 键 词：调制识别 时频图像 纹理特征 GABOR变换 

分 类 号：TN971[电子电信—信号与信息处理]