,Cumulant Based Harmonic Retrieval in Mixed Colored Gaussian and Non-Gaussian ARMA Noises

• 作 者: LI Shenghong ZHU Hongwen
• 机 构: Department of Electronic Engineering, Shanghai Jiao Tong University, Shanghai 200030, China
• 出 处: 中国电子杂志（英文版）
• 发表日期: 2001年2期
• 关键词: Autoregressive moving-average (ARMA) process Harmonic retrieval Phase coupling H

This paper studies the problem of retrieving one-dimensional real harmonic signals in presence of mixed colored Gaussian and non-Gaussian autoregressive moving-average (ARMA) noises, and proposes a new approach to harmonic retrieval. In the approach, Hilbert transform is first used to transform the real noisy observed data into their complex form; and then, some kind of fourth-order cumulant,which is defined particularly, is employed to identify the autoregressive (AR) parameters of the colored non-Gaussian ARMA noise model; after the real noisy observed data are filtered with the identified AR parameters, cumulant based methods can be used to compute the frequencies and the amplitudes of the harmonics. The proposed new approach can be applied to retrieve one-dimensional real harmonic signals in mixed colored Gaussian and non-Gaussian ARMA noises, no matter whether there is quadratic phase coupling or not in the harmonic signals and no matter whether the distribution of the colored non-Gaussian ARMA noise is symmetrical or not. Simulation examples are presented to demonstrate its effectiveness.