混合色噪声背景下的弱信号频率估计:基于互高阶累积量的矩和最小范围-TLS方法  

Weak Signal Frequency Estimation Embedded in Hybrid Colored Noise: Cross-High-Order Cumulant Based Moment and Minimum Norm-TLS Method

作　　者：马彦[1] 石要武[1] 戴逸松[1] 

机构地区：[1]吉林工业大学信息学院,长春130025

出　　处：《计量学报》2001年第3期234-238,共5页Acta Metrologica Sinica

基　　金：国家自然科学基金!(F0 10 2 0 2 6 9872 0 12 )

摘　　要：谱估计方法是用于噪声背景下测量微弱正弦信号的一种有效方法。但以往的谱估计都是采用互功率谱估计和高阶自谱估计的各种方法。本文在互功率谱估计和高阶统计量理论基础上 ,提出了互高阶谱估计方法。从理论上建立了基于互四阶累积量的Yule Walker方程 ,并在此基础上提出了互高阶谱估计的矩和最小范数 总体最小二乘 (TLS)方法。仿真结果表明 ,在混合色噪声背景下 ,这两种方法能够有效地抑制噪声 ,具有良好的频率估计性能。Spectral estimation method is an effective method in the case of parameter estimation of weak sinusoidal signal in noises. Spectral estimation always applies cross-spectral theory and high-order self-spectral theory up to now. In the paper, Yule-Walker equation based on cross-fourth-order cumulant has been developed firstly. Moment and minimum norm-TLS method are provided in this paper. Simulation results have shown that these two methods are effective to estimation weak sinusoidal signal frequency from hybrid colored noises and have good performance.

关 键 词：互四阶累积量 混合色噪声 频率估计 Yule-Walker方程 谱估计 弱信号检测 矩 最小范数 TLS方法 

分 类 号：TN911.23[电子电信—通信与信息系统]