华中科技大学学报（自然科学版）Issue(z1)：404-407,4.DOI:10.13245/j.hust.15S1096
基于H∞一致性滤波的分布式语音信号增强
Distributed speech signal enhancement based on H ∞-consensus filtering
摘要
Abstract
The speech signal enhancement problem in sensor networks was discussed,and a distribu-ted estimation approach was proposed based on the sparsity of signals on frequency domain.By desig-ning a node filter to fuse the measurements over the network,each sensor node got a satisfactory esti-mation.Firstly,a so-called l 1-regularized H ∞ filter was established by means of the classic H ∞ filte-ring theory and a pseudo-measurement equation.A high-pass consensus filter was embed into the l 1-regularized H ∞ filter in information form,and a distributed filtering algorithm was developed,which ensured that all node filters could reach a consensus on the estimates asymptotically and satisfied the prescribed H ∞ performance constraint.Finally,a numerical example was provided to demonstrate ef-fectiveness and applicability of the proposed method.关键词
分布式滤波/H ∞ 滤波/一致性滤波/传感器网络/语音信号增强Key words
distributed filtering/H ∞ filtering/consensus filtering/sensor network/speech signal en-hancement分类
信息技术与安全科学引用本文复制引用
于海洋,张汝波,刘冠群..基于H∞一致性滤波的分布式语音信号增强[J].华中科技大学学报（自然科学版）,2015,(z1):404-407,4.基金项目
辽宁省教育厅科学研究一般资助项目（L2014542）． （）
