基于矩阵重构的非相干分布源参数估计方法

针对现有非相干分布源参数估计算法计算量大等问题,提出了一种新的DOA和角度扩展估计算法.该算法利用空间频率近似模型下的非相干分布源协方差矩阵的结构特点.协方差矩阵可分为相位信息矩阵和幅度信息矩阵.对相位信息矩阵进行重构,对重构后的协方差矩阵利用MUSIC等传统方法即可实现DOA的估计;对幅度信息矩阵的各主次对角线上的元素均进行平滑得到幅度信息平滑向量,估计得到的DOA代入幅度信息中即可得到角度扩展的估计,从而实现非相干分布源DOA和角度扩展分离估计.计算机仿真验证了算法的性能.

Parameter estimator for incoherently distributed source based on matrix rebuild

To solve the high computing load of the parameter estimation algorithm for incoherently distributed source,a new DOA and angular spread estimation algorithm is proposed.This algorithm uses the structure character of the covariance matrix for incoherently distributed source approximated by a spatial frequency model.The covariance matrix can be separated into two matrixes rebuilt by phase information and amplitude information,respectively.The phase information matrix is rebuilt,thus the DOA could be estimated from the rebuilt phase information matrix by using traditional methods,e.g.MUSIC.The diagonal and sub-diagonals elements of the amplitude information matrix are averaged.The averaged vector which contains the amplitude information could be obtained.The angular spread could be estimated by substituting the estimated DOA into the amplitude information vector,thus realizes the separable estimation of DOA and angular spread for incoherently distributed source.Numerical examples illustrate the performance of the proposed method.