任意阵列二维测向的快速子空间算法

Fast Subspace Algorithm for 2-D Direction Finding Based on Arbitrary Antenna Array

为了对空间辐射源进行精确定位,建立了基于任意阵列对多目标源进行二维DOA估计的数学模型。将MUSIC算法推广到三维空间阵列可以对辐射源进行二维高精度测向,但由于其需要估计接收数据的协方差矩阵和进行特征分解,因而其计算量较大。利用多级维纳滤波器的前向递推获得信号子空间和噪声子空间,不需要估计协方差矩阵和对其进行特征分解,从而降低了MUSIC算法的计算量。将文中的方法应用于任意阵列的二维DOA估计中进行计算机仿真和实际侧向系统性能验证,实验结果均表明该方法达到了MU-SIC算法的性能,但与常规MUSIC算法相比降低了计算量。

In order to locate the radiant sources accurately, a mathematic model was established for arbitrary antenna array to estimate 2-D DOA of multiple sources, The MUSIC algorithm can be used to estimate the DOA of radiant sources precisely based on three-dimensional space antenna array. But it involves estimating the covariance matrix of observed data and its eigendecomposition, it has more computational complexity. Utilizing the forward recursion of the multistage wiener filter （MSWF）, we can get the signal and noise subspace, avoiding the estimation of the covariance matrix and eigendecomposition, and so the computational complexity is reduced. The method is used to estimate the 2-D DOA based on arbitrary space array, compute simulations and experimental results in practical system are given to demonstrate the efficiency of the proposed method.