一种基于ESPRIT的改进算法

An ESPRIT-Based improved algorithm

在目标波达方向估计的众多算法中,ESPRIT算法是一种运算速度较快、估值精度较高的常用算法。在该算法中,利用两个相同平移阵列接收信号的协方差矩阵进行广义特征值分解,进而求出目标波达方向的高分辨率估计,但是由于需要对协方差矩阵进行广义特征值分解运算,大大增加了计算的复杂度,因而不利于在工程应用中实现。本文提出了一种基于ESPRIT的改进算法,不需要对来波信号的协方差矩阵进行广义特征值分解,就可以实现对目标波达方向的高分辨估计,理论分析和仿真结果证明了该改进算法的有效性和可行性。

In the numerous algorithms of DOA estimation, ESPRIT algorithm has got traits of fairly fast calculation and high estimation accuracy, therefore it has been widely used in DOA estimation. This algorithm takes advantage of generalized eigenvalue decomposition of covariance matrices that consisted of two translating identical arrays. By taking this method,it can solve the problem of high-resolution DOA estimation. However, ESPRIT algorithm has greatly increased the complexity of calculating due to eigen-decomposition of covariance matrice formed by received signal＇ s arrays. It is inconvenient when the algorithm is used in the engineering applications. In this article, An ESPRIT-Based improved algorithm is presented. The proposed method is exploiting received signals from two translating identical arrays,like ESPRIT algorithm, but it needs not have generalized eigenvalue decomposition to resolve the problem of high-reso-, lution DOA estimation. Theoretical analysis and Simulation＇ s results verify that the proposed algorithm is effectiveness and feasibility.