基于压缩感知的高分辨DOA估计  

Compressive Sensing Based High Resolution DOA Estimation

作　　者：贺亚鹏[1] 李洪涛[1] 王克让[1] 朱晓华[1] 

机构地区：[1]南京理工大学电子工程与光电技术学院,南京210094

出　　处：《宇航学报》2011年第6期1344-1349,共6页Journal of Astronautics

基　　金：南京理工大学自主科研专项计划(2010ZYTS082)

摘　　要：研究基于压缩感知的多目标DOA估计问题。利用空间目标空域的稀疏性,提出一种新的DOA压缩感知模型,其感知矩阵满足约束等容(RIP)条件。提出一种基于奇异值分解的多测量矢量欠定系统正则化聚焦求解(SVD-RMFOCUSS)算法。该算法在一定程度上克服了压缩感知重构算法无法用于低信噪比情况缺陷,且具有较低的运算复杂度。计算机仿真表明该算法性能优于传统DOA估计算法,能够对任意相关性的信号进行有效DOA估计,具有更高的角度分辨力及估计精度。The multi-targets DOA estimation problem is investigated based on compressive sensing in this paper.Considering sparsity of the spatial targets,a novel DOA compressive sensing model is derived,in which sensing matrix satisfies the restricted isometry property（PIP）.A regularized multi-vectors focal undetermined system solver（RMFOCUSS） DOA algorithm based on singular value decomposition is proposed.The algorithm overcomes the defect that compressive sensing recovery algorithms can not be used in a low signal-to-noise ratio case,and has a lower computational complexity.Computer simulations show that the presented algorithm performs better than traditional DOA algorithms for signals with any coherence and offers higher angle resolution and estimation accuracy.

关 键 词：压缩感知 超分辨 波达方向 奇异值分解 多测量矢量欠定系统正则化聚焦求解算法 

分 类 号：TN911.7[电子电信—通信与信息系统]