基于对称压缩MUSIC的相干源方位估计方法

Azimuth Estimation Method for Coherent Source based on Symmetric Compression MUSIC

针对传统的多重子空间分类算法不适用于相干源条件且谱搜索计算量大的问题,提出一种基于对称压缩谱的互协方差空间平滑算法。首先,在传统的前后平滑基础上,最大限度利用数据互协方差矩阵来构建新的协方差矩阵,从而满足解相干。其次,对协方差矩阵进行特征值分解,利用对称压缩谱(MUSIC Symmetrical Compressed Spectrum,MSCS)思想,构造一种新的空间谱函数,即可在半谱范围内进行谱搜索得到完整的方位信息。所提算法解决了多重子空间分类算法不适用于相干源条件的问题,大大减少了算法空间谱搜索的计算量,一定程度上提高了算法的抗噪性能和分辨率,具有重要的现实意义。

Aiming at the problem that the traditional multiple subspace classification algorithms is not applicable to coherent source conditions and the amount of spectral search calculation is large, a cross-covariance spatial smoothing algorithm based on symmetric compression spectrum is proposed. Firstly, based on traditional front-to-back smoothing, the data covariance matrix is used to build up the new covariance matrix to the maximum, so as to satisfy de-coherence. Then, the eigenvalue decomposition is performed on the covariance matrix and a new spatial spectral function is constructed by using MSCS （MUSIC Symmetrical Compressed Spectrum） idea. Then spectrum search can be performed in the half spectrum, thus to obtain complete azimuth information. The proposed algorithm solves the problem that the multiple subspace classification algorithms is not applicable to coherent source conditions, and greatly reduces the computational complexity of the algorithm in spatial spectrum search. All this could improve, to some extent, the anti-noise performance and resolution of the algorithm, and thus has important practical significance.