English

应用科学学报 >

2013 , Vol. 31 >Issue 2: 165 - 169

DOI: https://doi.org/10.3969/j.issn.0255-8297.2013.02.010

信号与信息处理

非高斯噪声模型参数的最大似然快速估计

展开
  • 海军工程大学电子工程学院，武汉430033
蒋宇中，教授，博导，研究方向：通信相关技术、低频大气噪声和水下噪声建模等，E-mail：jiangyuzhong@tsinghua.org.cn

收稿日期: 2011-06-27

  修回日期: 2011-10-20

  网络出版日期: 2011-10-20

基金资助

国防科工委资助项目

收起

Fast Maximum Likelihood Estimation of Class A Model

Expand
  • College of Electronic Engineering, Naval University of Engineering, Wuhan 430033, China

Received date: 2011-06-27

  Revised date: 2011-10-20

  Online published: 2011-10-20

Fold

摘要

研究了Class A统计物理模型，对该模型的参数估计提出了一种最大似然方法，利用FFT的优势降低运算量，通过对观测数据进行分类计算以保证精度要求，并设计了初始值估计方案. 仿真实验表明，该方法精度高，迭代次数少，有较高的应用价值.

关键词： Class A噪声; 非高斯噪声; 最大似然估计; 快速傳里叶变换

本文引用格式

蒋宇中， 应文威， 刘月亮 . 非高斯噪声模型参数的最大似然快速估计[J]. 应用科学学报, 2013 , 31(2) : 165 -169 . DOI: 10.3969/j.issn.0255-8297.2013.02.010

Abstract

This paper investigates the Class A noise model, and proposes a method to determine parameters of the model based on maximum likelihood estimation. The method uses FFT to reduce computation complexity and enhance performance by calculating two data groups from the original observed data. A method for estimating initial values is also proposed. Simulation results show that the method has good performance with a small number of iterations, and therefore is suitable for practical applications.

Key words： class A noise; non-Gaussian noise; maximum likelihood estimation; fast Fourier transform (FFT)

参考文献

[1] ABRAHAM D A. Detection-threshold approximation for non-Gaussian backgrounds [J]. IEEE Journal of Oceanic Engineering, 2010, 35: 335-341.
[2] GOKEN C, GEZICI S, ARIKAN O, Optimal signaling and detector design for power-constrained binary communications systems over Non-Gaussian channels [J]. IEEE Transactions on Communications Letters, 2010, 14: 100-111.
[3] 蒋宇中，胡修林，张曙霞. 非高斯噪声中的Turbo码的性能改进研究 [J]. 应用科学学报，2006, 24(4): 336-340.
    JIANG Yuzhong, HU Xiulin, ZHANG Shuxia, Performance improvement of Turbo code in Non-Gaussian noise[J]. Journal of Applied Sciences, 2006, 24(4): 336-340.
[4] MIDDLETON D. Statistical-physical models of electromagnetic interference [J]. IEEE Trans.1977, EMC-19: 106-127.
[5] MIDDLETON D. Procedures for determining the parameters of the first-order canonical models of Class A and Class B electromagnetic interference [J]. IEEE Transactions on Electromagnetic Compatibility, 1979, EMC-21(3).
[6] ZABIN S M, POOR H V. Efficient identification of Non-Gaussian mixtures[J], IEEE Transactions on Communications, 2000, 48: 106-117.
[7] ZABIN S M, POOR H V. Effcient estimation of Class A noise parameters via the EM algorithms [J]. IEEE Transactions on Information Theory, 1991, 37: 60-72.
[8] JIANG Yuzhong. Bayesian estimation of Class A noise parameters with hidden channel states [C]// 2007 IEEE International Symposium on Power Line Communications and Its Applications, March, 2007: 26-28.
[9] MITTNIK S T, DOGANOGLU T, CHENYAO D. Computing the probability density function of the stable Paretian distribution[J]. Mathematical and Computer Modeling, 1999, 29: 235-240.
[10] JIANG Yuzhong. Identification of Class A Noise parameters via least square gradient method [C]// 2009 International Congress on Image and Signal Processing, Oct. 2009:17-19.
Options
文章导航

/

〈 〉