基于协方差、正性和l_1范数约束的迭代波达方向估计方法

A Recursive Approach to DOA estimation Based on Covariance、 Positive and l_1 Norm Constraint

本文通过将只能用于均匀线阵的ＰＨＤ方法推广到稀疏线阵，同时将正性约束引入ＦＯＣＵＳＳ方法，得到了一种基于协方差、正性和ｌ１范数约束进行协方差重建的迭代波达方向（ＤＯＡ）估计方法。该方法利用了ＭＵＳＩＣ方法忽略的反映阵列几何形状的协方差矩阵结构信息和ＤＯＡ估计的稀疏约束信息，不仅突破了信号源个数小于阵元数的限制，并具有提高ＤＯＡ估计性能的潜力。理论分析和仿真实验结果表明，这种迭代ＤＯＡ估计方法一般经过数次迭代就能获得稳定的高分辨率ＤＯＡ估计。

A recursive approach to DOA estimation has been developed by extending PHD method to sparse linear array and introducing positive constraint to FOCUSS method. Herein, we obtain the potential performance improvement of estimation and can estimate more sources by using the information about the structure of the covariance matrix and sparse character of DOA estimation, which was ignored by MUSIC method. Theory analysis and numerical results illustrate the convergence of this recursive DOA estimator with high resolution.