Communication channel equalization using wavelet network

被引：10

作者：

Pradhan, A. K.

Meher, S. K. ^{[1
]}

Routray, A.

机构：

[1] Natl Inst Sci & Technol, Dept Elect & Commun Engn, Berhampur 761008, Orissa, India

[2] Indian Inst Technol, Dept Elect Engn, Kharagpur 721302, W Bengal, India

来源：

DIGITAL SIGNAL PROCESSING | 2006年 / 16卷 / 04期

关键词：

channel equalization; neural network; wavelet network;

D O I：

10.1016/j.dsp.2005.06.001

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

Wavelet network (WN) based on wavelet decomposition principle is applied to channel equalization for both linear and non-linear channels. The WN is trained by extended Kalman filter (EKF) based recursive algorithm and is compared with EKF based multi-layered perceptron (MLP) and radial basis function neural network (RBFNN). Exhaustive simulation study reveals the superiority of the WN based equalizer in terms of bit error rate performance, compared to the above equalizer scheme. (C) 2005 Elsevier Inc. All rights reserved.

引用

页码：445 / 452

页数：8