一种基于改进CoSaMP的LFM雷达信号回波盲稀疏度估计方法  

An Evaluation Method of LFM Radar Echo Signal Unknown Sparse Degree Based on the Improved CoSaMP

作　　者：阮怀林[1] 王平[1] 

机构地区：[1]电子工程学院,合肥230037

出　　处：《电子对抗》2015年第1期14-20,共7页Electronic Warfare

摘　　要：由于压缩采样匹配追踪（CoSaMP）算法用于LFM雷达中，存在已知信号稀疏度的不足以及其迭代终止条件设置不够合理，文中对CoSaMP算法进行改进。在稀疏度未知时，由不同发射信号通过延时进而相加来构造冗余字典，改进CoSaMP算法依据残差之差的相对能量小于设定的停止门限来终止迭代过程，通过自适应调整可变步长逐步逼近信道的稀疏度，从而实现了重构LFM雷达回波信号。仿真结果表明，改进CoSaMP算法低信噪比时重构精度高于CoSaMP算法，当回波信号的距离分辨率匹配字典的距离分辨率，冗余字典结合改进CoSaMP算法可有效处理LFM雷达回波信号，具有广泛的应用价值和实用性。Because of the application of Compressive Sampling Matching Pursuit（CoSaMP） al- gorithm in LFM radar, the lack of kown signal sparse degree exists and the termination condi- tion of iteration is not set rationally. So the CoSaMP algorithm is improved in this paper. When the sparse degree is not known, the redundant dictionary is structured with the delay and addi- tion of the different signal. The improved CoSaMP algorithm end iteration process when the rel- ative energy of residual is less than the setting stopping threshold. The LFM radar echo signal is reconstructed with serf-adaptation adjusting variable step size to gain on the sparse degree of channel step by step. Simulation results show that the improved CoSaMP algorithm get higher reconstructing accuracy than the CoSaMP algorithm on the low signal noise ratio （SNR）. The LFM radar echo signal is handed efficiently with the redundant dictionary and the improved Co- SaMP algorithm when the distance of the echo signal matches the range resolution of the redun- dant dictionary. Therefore it has extensive value of application and extensive practical applica- bility.

关 键 词：压缩感知 线性调频 冗余字典 压缩采样匹配追踪算法 

分 类 号：TN957.51[电子电信—信号与信息处理]