摘要
针对传统极点提取算法在低信噪比时估计性能严重退化的缺点,该文从理论上推导了指数衰减正弦和信号在互相关处理后仍可表示为指数衰减正弦和信号的条件,定义了一种新的信号互相关函数估计方法,在此基础上提出了使用互相关消噪作为预处理的信号极点提取新方法.仿真实验表明,在相同精度的条件下文中提出的新方法较之改进KT法和基于四阶累量方法需要的信噪比降低了5dB.
In order to improve the estimation accuracy of the pole in low signal-noise-ra tio,it is theoretically analyzed that the signals composed of the exponentiall y damped sinusoids can be still expressed as exponentially damped sinusoids after the cross-correlation.A new cross-correlation estimator is proposed,and then a novel method of the poles extraction based on the cross-correlation is given. The simulations show that in the same estimation accuracy SNR required by the no vel method is 5dB lower than that of the modified Kumaresan-Tufts method or fou rth-order cumulants based method.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2005年第6期1015-1018,共4页
Acta Electronica Sinica
关键词
指数衰减正弦
瓦相关估计
极点提取
exponentially damped sinusoids
cross-correlation estimator
poles extracting