基于稀疏成份分析的高分辨成像算法  

High Resolution Imaging Algorithm Based on Sparse Component Analysis

作　　者：王良[1] 尚朝轩[1] 何强[1] 韩壮志[1] 

机构地区：[1]解放军军械工程学院电子与光学工程系,河北石家庄050003

出　　处：《探测与控制学报》2013年第5期31-36,共6页Journal of Detection & Control

摘　　要：Costas编码信号具有图钉形的模糊函数,不存在速度-距离耦合现象,能同时实现距离和速度测量,但其测距和测速精度受到带宽和时宽限制,并且比较高的模糊旁瓣会造成大目标对小目标的遮蔽效应。针对这一特性,根据目标在距离-速度二维稀疏的先验信息,提出一种高分辨的二维成像方法。该方法采用波形熵作为度量函数,构造冗余词典,将乘子法用于推广的正则化FOCUSS(Focal underdetermined system solver)算法,求解约束极小化问题的最优解,实现回波信号的稀疏成份分析。仿真结果表明,该方法实现了距离-速度二维的高分辨成像,并且避免了大目标对小目标的遮挡效应。Although the Costas-eoded hopping signal has thumbtack shaped ambiguity function and no velocity- range coupling, the resolution in range and velocity is constrained by the bandwidth and time duration, and it is possible that large targets will mask the small ones. According to these characters and the prior information that the target was sparse in range-velocity domain, a super-resolution algorithm was proposed. Using waveform en- tropy as the sparsity measure function, a redundancy dictionary was constructed by using the method of La- grange multipliers, the generalized regularized FOCUSS （focal underdetermined system solver） algorithm was improved to get the optimal solution of the constrained minimization problem, and the sparse component analysis was realized. The Super-resolution in range-velocity domain is verified via simulation, and the large targets masking on the small ones was solved.

关 键 词：COSTAS编码 释疏成份分析 波形熵 

分 类 号：TN957[电子电信—信号与信息处理]