基于非均匀时延的频率估计改进算法

ImprovedAlgorithm for Frequency Estimation Based on Non-uniform Delay

为了减少算法运算量,将多级维纳滤波器引入非均匀时延频率估计中,对原有算法进行改进.改进算法通过简单的乘加运算得到噪声子空间,无需对观测数据协方差矩阵进行特征根分解,因此也更加易于硬件实现;结合一维MUSIC高分辨算法完成对信号频率的参数估计.仿真结果表明:在信噪比大于0dB时,改进算法与原有算法性能相当.

This paper is to introduce the multi-stage Wiener filter（MSWF） to non-uniform delay frequency estimation for reducing the computation of the original algorithm,improving the original one.By the simple multi-plication and addition operation in the improved algorithm,it obtains the noise sub-space,and is no need to do any eigenvalue decomposition of the covariance matrix of the observed data,thus to have an easy hardware implemen-tation.Finally,it carries out the parameter estimation of signal frequency in terms of one-dimensional MUSIC high resolution algorithm.Simulation results show that the performance of the improved algorithm matches to that of the original one when the SNR is above 0 dB.