基于时频图像处理的多分量LFM信号分离  

Separation of multi-component LFM signals based on image processing in time frequency domain

作　　者：刘歌[1] 张国毅[1] 胡鑫磊[1] 王晓峰[2] 

机构地区：[1]空军航空大学信息对抗系,吉林长春130022 [2]海军航空工程学院,山东烟台264001

出　　处：《航天电子对抗》2015年第5期46-49,59,共5页Aerospace Electronic Warfare

摘　　要：针对传统时频分析方法对时频域交叠的多分量线性调频(LFM)信号分离不足的问题,提出一种基于图像处理技术的分离方法。该方法首先计算多分量LFM信号的平滑伪Wigner-Ville分布(SPWVD),并通过Radon变换估计出各分量的调频斜率。然后利用形态学处理方法处理SPWVD二值图像,实现多分量LFM信号的粗分离,最后采用连通区域标记法实现信号分量的细分离。仿真实验表明,该方法能够在低信噪比条件下对多分量LFM信号进行有效分离。Conventional time-frequency analysis methods are exposed to some limitations in separating multi-component LFM signals. A method for time-frequency separation of multi component LFM signals is proposed based on image processing technology. First, the method applies the Smooth Pseudo WignevVille distribution （SPWVD） to multi-component LFM signals, and estimates the frequency modulation slope of the LFM by performing Radon transform. Then the time-frequency image is converted into binary image which will be processed by the morphology to separate multi-component LFM signals roughly. Finally, the method of connected component labeling is used to realize the exact separation of signal components. Simulation results show that the proposed approach is efficient in the condition of low SNR.

关 键 词：电子侦察 多分量LFM信号 时频分布 图像处理 信号分离 

分 类 号：TN911.73[电子电信—通信与信息系统]