Modifier formula on mean square convergence of LMS algorithm

In describing the mean square convergence of the LMS algorithm, the update formula based on independence assumption will bring explicit errors, especially when step-size is large. In this paper, a modifier formula that describes the convergence well, is proposed. The simulations support the proposed...

Full description

Saved in:
Bibliographic Details
Published inElectronics letters Vol. 38; no. 19; p. 1
Main Authors Gu, Yuantao, Tang, Kun, Cui, Huijuan, Du, Wen
Format Journal Article
LanguageEnglish
Published Stevenage John Wiley & Sons, Inc 12.09.2002
Subjects
Algorithms
Computer simulation
Convergence
Errors
Mean square values
Online AccessGet full text
ISSN0013-5194
1350-911X

Cover

More Information
Summary:In describing the mean square convergence of the LMS algorithm, the update formula based on independence assumption will bring explicit errors, especially when step-size is large. In this paper, a modifier formula that describes the convergence well, is proposed. The simulations support the proposed formula in different conditions.
Bibliography:SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 14
ObjectType-Article-1
ObjectType-Feature-2
content type line 23
ISSN:0013-5194
1350-911X