摘要
针对在低信噪比条件下雷达辐射源信号识别率低的问题,为了提高雷达信号识别率,提出了一种采用多特征融合的脉内调制方式识别方法。首先对信号进行Choi-Williams变换,得到时频图后实施降噪处理;然后利用奇异值分解(SVD)和线性鉴别分析(LDA)两种方法提取其时频图特征值,并进行特征融合,最后选择用最小距离准则进行分类判别。仿真选用6种常见的雷达辐射源信号,仿真结果表明在0dB的低信噪比条件下,上述方法的平均识别率在90%以上。最后将特征融合前后的识别效果进行对比,仿真结果验证了融合算法的优越性,证明可为雷达信号优化识别提供依据。
To solve the problem of low rate in radar emitter signal recognition under low SNR, a new approach u- sing fusion of features for recognition of intra-pulse modulation is proposed. At first, the time-frequency images of radar emitter signals are obtained by Choi-Williams transform, and then the noise reduction of these images is pro- cessed. After that, the methods of SVD and LDA are used to extract the features of time-frequency images and the features fusion is implemented afterwards. At last, the minimum distance principle is used for classification and dis- crimination. Six kinds of common radar signals are used in simulations. The results show that the average recognition rate can reach 90% or more when the SNR is as low as 0dB. The recognition effect after fusion is also compared with that without using it. The superiority of the fusion algorithm is proved by simulation results.
出处
《计算机仿真》
CSCD
北大核心
2016年第3期18-22,共5页
Computer Simulation
基金
国家自然科学基金资助项目(61372166)
陕西省自然科学基础研究计划资助项目(2014JM8308)
