改进MUSIC算法的DOA估计眭能研究

在众多的波达方向（DOA）估计算法中，以MUSIC算法为代表的基于特征结构分解方法因具有较高的分辨性能而受到研究者的青睐。但是当信号源完全相干时，其性能会显著变坏。改进了MUSIC算法．通过构造转换矩阵，对协方差矩阵进行变换，将其转换成实矩阵再进行特征值分解，实现了MUSIC算法解相干。计算机仿真证明了该方法的有效性。

Research on DOA Performance of Improved MUSIC

In a number of DOA（Direction-Of-Arrival） estimation algorithms, the method based on feature structure decomposition with MUSIC as a representative has attracted significant coverage from researchers with high resolution capability. However, its performance becomes worse with coherent sources. The MUSIC is improved with constructing conversion matrix, transforming covariance matrix to real matrix and eigendecomposition. The ability of decoherence with MUSIC is realized. The simulation proves its availability.