脉冲噪声环境下基于相关熵的AR模型α谱估计方法

在高脉冲水平条件下,基于分数低阶统计量的α谱估计方法对脉冲噪声变得敏感,并且分数阶参数值的选取影响估计精度,其性能会出现不同程度的下降.根据信号的相关熵函数对非高斯噪声不敏感的特点,提出了一种基于相关熵的模型α谱估计方法.仿真结果显示,在相同混合信噪比条件下,所提方法具有无需预设分数阶参数值和顽健性强的特点,特别在高脉冲水平噪声下相比其他方法频率分辨能力更具优势....

Full description

Saved in:
Bibliographic Details
Published in电子技术应用 Vol. 40; no. 11; pp. 105 - 108
Main Author 李静威 全厚德 崔佩璋
Format Journal Article
LanguageChinese
Published 军械工程学院信息工程系,河北石家庄,050003 2014
Subjects
AR模型谱估计
分数低阶协方差
分数低阶矩
相关熵函数
fractional lower order moment
fractional lower order covariance
分数低阶矩
AR模型谱估计
correntropy function
AR model α spectrum estimation
相关熵函数
分数低阶协方差
Online AccessGet full text
ISSN0258-7998

Cover

More Information
Summary:在高脉冲水平条件下,基于分数低阶统计量的α谱估计方法对脉冲噪声变得敏感,并且分数阶参数值的选取影响估计精度,其性能会出现不同程度的下降.根据信号的相关熵函数对非高斯噪声不敏感的特点,提出了一种基于相关熵的模型α谱估计方法.仿真结果显示,在相同混合信噪比条件下,所提方法具有无需预设分数阶参数值和顽健性强的特点,特别在高脉冲水平噪声下相比其他方法频率分辨能力更具优势.
Bibliography:Li Jingwei, Quan Houde, Cui Peizhang （Information Engineering Department, Ordnance Engineering College, Shijiazhuang 050003, China）
correntropy function;AR model α spectrum estimation;fractional lower order moment;fractional lower order covariance
11-2305/TN
Fractional Lower Order Statistics （FLOS） based methods for α spectrum estimation are sensitive to the impulsive noise in big impulse noise environments,the selection of fractional lower order parameters affects estimation accuracy,and the performances of these methods are with different degrees of decline.This paper proposes the correntropy based method for AR model α spectrum estimation according to the correntropy function of signal insensitive to the non-Gaussian noise.Simulation results show that the new method is without default value of fractional lower order and very robust,and its frequency resolution has great advantages as compared with other methods under the same Mixed Signal to Noise Ratio（MSNR）.
ISSN:0258-7998