A constant modulus algorithm for blind equalization in α-stable noise

Channel noise is often assumed to be Gaussian in most of the existing channel equalization algorithms. The performance of these algorithms will degrade seriously when the noise is non-Gaussian. This paper deals with the problem of blind channel equalization in impulsive noise environment that is mod...

Full description

Saved in:
Bibliographic Details
Published in:Applied acoustics 2010-07, Vol.71 (7), p.653-660
Main Authors: ZHANG, Yin-Bing, ZHAO, Jun-Wei, GUO, Ye-Cai, LI, Jin-Ming
Format: Article
Language:English
Subjects:
Acoustics
Blind equalization
Constant modulus
Error-constrained
Exact sciences and technology
Fundamental areas of phenomenology (including applications)
Impulsive noise
Physics
Underwater sound
α-Stable noise
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Channel noise is often assumed to be Gaussian in most of the existing channel equalization algorithms. The performance of these algorithms will degrade seriously when the noise is non-Gaussian. This paper deals with the problem of blind channel equalization in impulsive noise environment that is modeled as α-stable process. A modified adaptive error-constrained constant modulus algorithm (MAECCMA) is proposed by soft-limiting the amplitude of the equalizer input and transforming the error signal of the original adaptive error-constrained constant modulus algorithm (AECCMA) nonlinearly to suppress the influence of α-stable noise. Computer simulation results of two underwater acoustic channels show that, MAECCMA has almost the same performance as AECCMA and they both have faster convergence rate than constant modulus algorithm (CMA) and normalized least mean absolute deviation (NLMAD) algorithm in Gaussian noise, while MAECCMA provides the best performance of those four algorithms in α-stable noise.
ISSN:0003-682X
1872-910X
DOI:10.1016/j.apacoust.2010.02.007