STAP中的矩阵求逆问题研究

针对空时自适应处理（STAP）中的采样协方差矩阵求逆问题，利用其为一正定Hermite阵这一特点，提出了一种利用Hermite阵对称性分块求逆的矩阵求逆新算法，并对该算法的运算量进行了定量分析和比较。从分析结果和仿真实验可以看出，该算法计算量小，存储量小，易于工程实现，相对于实际工程中较常用的Gauss消元法具有较明显的性能改善。

Study on Matrix Inversion for STAP

Based on the characteristic that the sample covariance matrix in the space-time adaptive processing（STAP） is Hermite and positive definite, this paper presents a new algorithm of matrix block-inversion. The computation cost of the new algorithm is analyzed and compared with the routine algorithms. As can be seen from the analysis results and computer simulations, the new algorithm is of less computation, storage smaller, and feasible, which achieves an obvious performance improvement in comparison with the Gauss elimination which is commonly employed in applications.