基于压缩感知的DOA估计稀疏化模型与性能分析

发布时间：2018-01-06 09:12

本文关键词：基于压缩感知的DOA估计稀疏化模型与性能分析　出处：《电子与信息学报》2014年03期 　论文类型：期刊论文

【摘要】：利用压缩感知理论解决阵列信号到达角(DOA)估计问题,具有对快拍数据量要求低、可处理相关源等优点。将压缩感知理论应用于信源DOA估计的一个关键问题是建立信源信号的稀疏化模型。该文在均匀线阵模型下系统分析了角度划分对DOA估计稀疏重构性能的影响,从对相关性的分析出发给出了信号的最优稀疏化模型。分析结果表明在实际应用中基于信源信号等正弦空间稀疏化的重构模型是最优的。实验对比了新的稀疏化模型与传统的等角度划分方式得到的流形矩阵的可重构性能,并进行了关于信号重构和信源DOA估计的详细实验分析,验证了所提模型的优越性。
[Abstract]:The theory of compressed sensing is used to solve the problem of DOA estimation of array signals, which has a low requirement for the amount of rapid-beat data. A key problem in applying compressed sensing theory to source DOA estimation is to establish a sparse source signal model. In this paper, angle partitioning for do is systematically analyzed under uniform linear array model. A estimates the effect of sparse reconstruction performance. Based on the analysis of correlation, the optimal sparse model of signal is given. The results show that the reconstruction model based on the sparsity of sinusoidal space such as source signal is optimal in practical application. The new sparse model is compared experimentally. The reconfigurable properties of the manifold matrix obtained by the model and the traditional equal-angle partition method. A detailed experimental analysis on signal reconstruction and source DOA estimation is carried out to verify the superiority of the proposed model.
【作者单位】：国防科技大学理学院;
【基金】：国家自然科学基金(61002024,61072142,61102169,61201332,61205190,61271437) 国防科学技术大学科研计划项目(JC11-02-03)资助课题
【分类号】：TN911.7
【正文快照】： 1引言阵列信号到达角(Direction-Of-Arrival,DOA)估计是阵列信号处理领域的关键问题,在雷达、通讯、地震等众多领域有广泛应用。空间谱估计技术的发展实现了目标DOA的超分辨估计。经典的空间谱估计算法包括MUSIC和ESPRIT算法,它们主要是根据阵列接收信号的统计特性来估计目标

【共引文献】