一种无需信源数目的方位估计方法

Direction of arrival estimation algorithm without knowing source number

针对子空间类波达角估计算法虽然性能可以达到超分辨,但当信源个数估计错误时,从而使得子空间划分错误将导致子空间类算法性能急剧下降的问题,本文提出一种无需信源数的超分辨波达角估计算法。该方法利用特定协方差矩阵的最小特征值对应的向量的正交性构建空间谱,因此算法无需已知信源数目,并讨论了所提算法下参数的边界条件以及选参方法。本文算法实现简单,与无需信源个数的最小方差无畸变响应算法相比拥有更高的分辨率以及更好的估计精度。通过计算机仿真实验分析验证了算法的可行性。

Although the subspace-based direction-of-arrival(DOA)estimation algorithm can achieve super-resolution,its performance decreases sharply when there is an incorrect subspace division due to the incorrect estimation of the source number.In this paper,we propose a super-resolution DOA estimation algorithm that solves this problem.This algorithm mainly uses the orthogonality of the minimum eigenvector of a specific covariance matrix to construct a spatial spectrum,so it does not require a known source number.The parameter selection method and boundary conditions of this algorithm are discussed in depth.The proposed algorithm is simple to implement.It also has higher resolution and estimation accuracy than the minimum variance undistorted response algorithm,which does not require the source number.The results obtained in computer simulation experiments verify the feasibility of the proposed algorithm.