基采用时频区域检测的雷达信号盲分选算法  

Blind Sorting Algorithm of Radar Signals Using Time-frequency Region Detection

作　　者：任枫轩[1] 王忠勇[2] REN Feng-xuan;WANG Zhong-yong(Department of Electrical Engineering, Henan Polytechnic, Zhengzhou 450046, China;School of Information Engineering, Zhengzhou University, Zhengzhou 450001, China)

机构地区：[1]河南职业技术学院电气工程系,河南郑州450046 [2]郑州大学信息工程学院,河南郑州450001

出　　处：《数学的实践与认识》2019年第12期166-173,共8页Mathematics in Practice and Theory

基　　金：河南省科技攻关项目(182102210085)

摘　　要：为了解决欠定条件下密集雷达信号分选问题,提高雷达信号盲分选算法精度,提出了一种采用时频区域检测的雷达信号盲分选算法,首先利用短时傅立叶变换将混合信号映射至时频平面进行处理,然后通过聚类算法估计出混合矩阵,从而反解出源信号矩阵,进而估计出每一雷达源信号,能够在信号的时频平面投影相交且欠定的条件下,实现信号分选功能.仿真实验结果表明:提出的算法相比于传统的类MUSIC算法及其衍生的相关改进算法具有更高的分选精度和算法收敛性,且估计得到的源信号时域波形更优,体现了其有效性和优越性.In order to solve the dense radar signal sorting problem under underdetermined condition, a radar signal blind sorting algorithm using time-frequency region det ection is proposed to improve the accuracy. Firstly, the mixed signal is mapped to the time-frequency region for processing by the short time Fourier transformation, then the hybrid matrix is estimated throgh clustering algorithm, so the source signal matrix is inversely solved, and each radar source signal is estimated. The proposed method can achieve the function of signal sorting when the signal is projected and intersected in the time-frequency region, under underdetermined conditions. The simulation results show that the proposed algorithm has a higher sorting accuracy and algorithm convergence comparing with the traditional MUSIC algorithm and its improved algorithms, and estimates better time-domain waveform of the source signal, which reflects the effectiveness and superiority.

关 键 词：欠定条件 时频区域检测 短时傅里叶变换 雷达信号 盲分选 

分 类 号：TN957.51[电子电信—信号与信息处理]