一种高分辨率谱估计与数据外推的快速算法

A Fast Algorithm of High-Resolution Spectral Estimation and Data Sequences Extrapolation

利用信号频谱幅度分布的先验信息 ,对短时数据序列信号提出了一种实现高分辨率谱估计和数据序列外推的快速算法。该算法利用了算子矩阵的特殊结构 ,采用简单、快速的矩阵求逆算法 ,使算法的运算量和数据存储容量大幅度下降。给出了算法在高分辨率谱估计及数据序列外推中的应用实例和仿真结果 ,证实了该算法的优良性能。

In this paper, a fast algorithm of high-resolution spectral estimation and data sequences extrapolation for short-time data sequences is proposed by using the prior distribution of spectral amplitudes. This algorithm which computes the inversion of operator matrix is simple and fast based on the special structure of operator matrix. The computational load and the number of data memory units of the algorithm can be efficiently reduced. The paper also gives out the application examples and computer simulation results that demonstrate the excellent performance of high-resolution spectral estimation and data sequences extrapolation