一种非均匀噪声下的DOA估计算法

A DOA Estimation Algorithm in Non-Uniform Noise

通用协方差差分算法用来实现对空间非均匀噪声环境下相干信号的波达方向(DOA)估计,该算法可以完全消除空间非均匀噪声,且适用于低信噪比环境,但该算法的DOA估计结果存在伪峰。针对这一问题,提出了一种改进的算法。改进算法通过对通用协方差差分(GCD)算法的信号协方差矩阵进行变换,再用特征分解的方法得到信号的DOA估计值。改进的算法可以完全消除伪峰,理论分析和仿真实验验证了改进算法的有效性。

Generalized covariance differencing （GCD） algorithm is proposed for direction of arrival （DOA） estimation in the presence of coherent sources and unknown non-uniform noise. It can fully elim- inate spatially non-uniform noise and fit for low signal to noise ratio （SNR） environments. However, the DOA result of GCD algorithm has phantom peaks. To resolve this problem, an improved algorithm is pro- posed. A new source eovarianee matrix can be obtained by the transformation of the source covariance matrix of GCD algorithm. Then, actual DOA of the source is achieved by eigenvalue decomposition of the new source eovariance. The improved algorithm can totally eliminate the phantom peaks. The computer simulation testifies the validity of the improved algorithm.