HMM信道建模中齐次马氏链极限分布的矩阵分析

北京邮电大学学报

HMM信道建模中齐次马氏链极限分布的矩阵分析

北京邮电大学电信工程学院 , 北京 100876

基金资助:
国家自然科学基金资助项目 ( 60172052, 90104018) ; 863国家重大项目: 2001AA123016

On Matrix Eigenvalue Analysis of Homogeneous Markov Chains in Hidden Markov Channel Modeling

Telecommunication Engineering School, Beijing University of Posts and Telecommunications, Beijing 100876, China

Supported by:

摘要/Abstract

摘要： HMM信道建模中频繁用到状态转移概率矩阵 P 平稳分布求解. 本文首次通过对马氏链的转移概率矩阵进行矩阵特征值分析 ,给出了马氏链遍历性的一个充要判别定理. 最后利用这个充要条件 ,给出了平稳马氏链极限分布的一种矩阵特征值分解求法 .

关键词: 隐马尔可夫过程, 矩阵分析 , 特征值分解, 信道模型

Abstract: A necessary and sufficient condition is derived to determine the ergodicity of a Markov chain by using eigenvalue analysis of a square matrix, thus a novel approach is proposed to solve stationary-distribution problem in HMM-channel modeling, which is more straightforward and simpler to implement than conventional methods.

Key words: hidden markov process, matrix analysis, eigenvalue decomposition, channel modeling

中图分类号:

TN911. 1

邝育军,　侯晓林,　郝建军,　乐光新 . HMM信道建模中齐次马氏链极限分布的矩阵分析[J]. 北京邮电大学学报.

KUANG Yu-jun,　 HOU Xiao-lin,　 HAO Jian-jun,　 YUE Guang-xin . On Matrix Eigenvalue Analysis of Homogeneous Markov Chains in Hidden Markov Channel Modeling [J]. JOURNAL OF BEIJING UNIVERSITY OF POSTS AND TELECOM.

[1]	董秀则张茹钮心忻. MB隐写分析算法[J]. 北京邮电大学学报, 2012, 35(2): 99-103.
[2]	高向川,李立华，张平，宋磊，王朝炜. GEVD辅助的MMSE干扰抑制下行多用户MIMO系统[J]. 北京邮电大学学报, 2011, 34(1): 85-89.