摘要
基于对超定Hankel阵的SVD,提出一种估计隐含于未知有色噪声背景中正弦信号频率的方法。它适用于从观测数据出发直接估计正弦波频率。对背景噪声模型不做任何先验假定,只要它是任何一种非确定性的广义平稳过程,即它满足T_n(k)=0;k→∞即可,因而它适用于广泛一类有色噪声污染正弦信号的谱估计问题。
Based on SVD (singular-value decomposition) of an .overdetermined extended-order Hankel matrix, an approach to estimate frequencies of sinusoids in cglored noise is presented. It is applicable to spectral estimations of sinusoids in wide sense stationary(WSS) noise with condition r.(k) = 0, k - ∞, where {r.(k)} represent the autocorrelations of the noise.
出处
《四川大学学报(自然科学版)》
CAS
CSCD
1993年第3期348-352,共5页
Journal of Sichuan University(Natural Science Edition)
关键词
奇异值分解
正弦波
频率
谱估计
spectral estimation
SVD
sinusoidal frequency estimation.