一种基于预估计的稀疏表示DOA估计快速算法  

Fast DOA estimation algorithm with sparse representation based on pre-estimation

作　　者：张兴良[1] 王可人[1] 刘俊彤[1] 

机构地区：[1]电子工程学院通信对抗系,合肥230037

出　　处：《计算机应用研究》2014年第12期3596-3598,共3页Application Research of Computers

基　　金：国家自然科学基金资助项目(61171170)

摘　　要：稀疏表示DOA估计算法具有很高的分辨率,但是其计算量较大。为此,提出一种基于预估计的稀疏表示DOA估计快速算法。首先,利用阵列协方差矩阵特征向量的性质,通过DFT对入射信号进行DOA预估计;其次,利用预估计结果降低过完备字典的长度;最后,进行稀疏分解得到DOA估计的精确结果。由于采用的字典长度较短,该算法的计算量非常小。仿真实验结果表明,该算法具有较高的成功概率、较快的估计速度和较低的估计误差。因此,提出的算法在工程实际中应用前景广阔。Though the DOA estimation with sparse representation has high resolution,its computational load is too much. To settle this problem,this paper proposed a fast DOA estimation algorithm with sparse representation based on pre-estimation.First,with the feature of eigenvector of covariance matrix,it preliminary estimated the DOAs of incident signal by DFT. Then,using the result of pre-estimation,it reduced the length of the over-complete dictionary. Finally,it exactly estimated the DOAs by performing sparse decomposition. The computational load of the proposed algorithm is very small for the short dictionary length. The results of simulation experiments show that the proposed algorithm has high probability of success,fast estimation speed,and low estimation error. Therefore,there is much application prospect of the proposed algorithm in engineering practice.

关 键 词：阵列信号 波达方向 稀疏表示 特征矢量 预估计 

分 类 号：TN911.23[电子电信—通信与信息系统] TP301.6[电子电信—信息与通信工程]