基于LMD构建虚拟接收阵列的欠定语音盲提取  

Undetermined speech extraction of virtual receiver array based on LMD

作　　者：姜兴宗 王尔馥[2] JIANG Xing-Zong;WANG Er-Fu(Heilongjiang Moble Qiqihar Company Co.,Ltd.!Qiqihar 161000,Heilongjiang,China;chool of Electronic Engir^eringy Heilongjiang University,Harbin 150080,Chin)

机构地区：[1]黑龙江移动齐齐哈尔分公司,黑龙江齐齐哈尔161000 [2]黑龙江大学电子工程学院,哈尔滨150080

出　　处：《黑龙江大学工程学报》2018年第1期60-63,70,共5页Journal of Engineering of Heilongjiang University

基　　金：国家自然科学基金资助项目(61571181)

摘　　要：针对通信系统中存在的接收端天线个数少于发送端个数的欠定系统问题,提出利用局部均值分解(LMD)构建虚拟接收阵列,补充缺失的天线阵元以实现欠定系统转正定系统,在不同维度的混沌系统中实现数字语音信号的盲提取,以此证明了此算法的有效性、普适性以及可靠性。The problem that tlie receiving end antenna number is less than the number of transmitting end communication system is analyzed,we propose that using Local Mean Decomposition(LMD)to construct virtual receiver arrays and to supplement the missing antenna array element!realizes the converting of an underdetermined system to a positive definite system.The blind extraction of digital speech signals can be carried out in different dimensional chaotic systems.Thus,the validity and universality of the algorithm and the reliability of digital technology in the background of noise are proved.

关 键 词：局部均值分解 时频分析 提取 混沌 

分 类 号：TN911.7[电子电信—通信与信息系统]