New variable step-sizes minimizing mean-square deviation for the lms-type algorithms

Shengkui Zhao, Douglas L Jones, Suiyang Khoo, Zhihong Man

Research output: Contribution to journalArticle

Abstract

The least-mean-square-type (LMS-type) algorithms are known as simple and effective adaptation algorithms. However, the LMS-type algorithms have a trade-off between the convergence rate and steady-state performance. In this paper, we investigate a new variable step-size approach to achieve fast convergence rate and low steady-state misadjustment. By approximating the optimal step-size that minimizes the mean-square deviation, we derive variable step-sizes for both the time-domain normalized LMS (NLMS) algorithm and the transform-domain LMS (TDLMS) algorithm. The proposed variable step-sizes are simple quotient forms of the filtered versions of the quadratic error and very effective for the NLMS and TDLMS algorithms. The computer simulations are demonstrated in the framework of adaptive system modeling. Superior performance is obtained compared to the existing popular variable step-size approaches of the NLMS and TDLMS algorithms.

Original languageEnglish (US)
Pages (from-to)2251-2265
Number of pages15
JournalCircuits, Systems, and Signal Processing
Volume33
Issue number7
DOIs
StatePublished - Jul 2014

Fingerprint

Variable Step Size
Mean Square
Deviation
Least Mean Square
Mathematical transformations
Transform
Convergence Rate
Adaptive systems
Adaptive Systems
System Modeling
Time Domain
Quotient
Computer Simulation
Trade-offs
Minimise
Computer simulation

Keywords

  • Convergence
  • Discrete transforms
  • Least mean square algorithms
  • Mean-square error

ASJC Scopus subject areas

  • Signal Processing
  • Applied Mathematics

Cite this

New variable step-sizes minimizing mean-square deviation for the lms-type algorithms. / Zhao, Shengkui; Jones, Douglas L; Khoo, Suiyang; Man, Zhihong.

In: Circuits, Systems, and Signal Processing, Vol. 33, No. 7, 07.2014, p. 2251-2265.

Research output: Contribution to journalArticle

@article{fde6421973cd4672ba64d1863e2fcd26,
title = "New variable step-sizes minimizing mean-square deviation for the lms-type algorithms",
abstract = "The least-mean-square-type (LMS-type) algorithms are known as simple and effective adaptation algorithms. However, the LMS-type algorithms have a trade-off between the convergence rate and steady-state performance. In this paper, we investigate a new variable step-size approach to achieve fast convergence rate and low steady-state misadjustment. By approximating the optimal step-size that minimizes the mean-square deviation, we derive variable step-sizes for both the time-domain normalized LMS (NLMS) algorithm and the transform-domain LMS (TDLMS) algorithm. The proposed variable step-sizes are simple quotient forms of the filtered versions of the quadratic error and very effective for the NLMS and TDLMS algorithms. The computer simulations are demonstrated in the framework of adaptive system modeling. Superior performance is obtained compared to the existing popular variable step-size approaches of the NLMS and TDLMS algorithms.",
keywords = "Convergence, Discrete transforms, Least mean square algorithms, Mean-square error",
author = "Shengkui Zhao and Jones, {Douglas L} and Suiyang Khoo and Zhihong Man",
year = "2014",
month = "7",
doi = "10.1007/s00034-014-9744-2",
language = "English (US)",
volume = "33",
pages = "2251--2265",
journal = "Circuits, Systems, and Signal Processing",
issn = "0278-081X",
publisher = "Birkhause Boston",
number = "7",

}

TY - JOUR

T1 - New variable step-sizes minimizing mean-square deviation for the lms-type algorithms

AU - Zhao, Shengkui

AU - Jones, Douglas L

AU - Khoo, Suiyang

AU - Man, Zhihong

PY - 2014/7

Y1 - 2014/7

N2 - The least-mean-square-type (LMS-type) algorithms are known as simple and effective adaptation algorithms. However, the LMS-type algorithms have a trade-off between the convergence rate and steady-state performance. In this paper, we investigate a new variable step-size approach to achieve fast convergence rate and low steady-state misadjustment. By approximating the optimal step-size that minimizes the mean-square deviation, we derive variable step-sizes for both the time-domain normalized LMS (NLMS) algorithm and the transform-domain LMS (TDLMS) algorithm. The proposed variable step-sizes are simple quotient forms of the filtered versions of the quadratic error and very effective for the NLMS and TDLMS algorithms. The computer simulations are demonstrated in the framework of adaptive system modeling. Superior performance is obtained compared to the existing popular variable step-size approaches of the NLMS and TDLMS algorithms.

AB - The least-mean-square-type (LMS-type) algorithms are known as simple and effective adaptation algorithms. However, the LMS-type algorithms have a trade-off between the convergence rate and steady-state performance. In this paper, we investigate a new variable step-size approach to achieve fast convergence rate and low steady-state misadjustment. By approximating the optimal step-size that minimizes the mean-square deviation, we derive variable step-sizes for both the time-domain normalized LMS (NLMS) algorithm and the transform-domain LMS (TDLMS) algorithm. The proposed variable step-sizes are simple quotient forms of the filtered versions of the quadratic error and very effective for the NLMS and TDLMS algorithms. The computer simulations are demonstrated in the framework of adaptive system modeling. Superior performance is obtained compared to the existing popular variable step-size approaches of the NLMS and TDLMS algorithms.

KW - Convergence

KW - Discrete transforms

KW - Least mean square algorithms

KW - Mean-square error

UR - http://www.scopus.com/inward/record.url?scp=84902828135&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84902828135&partnerID=8YFLogxK

U2 - 10.1007/s00034-014-9744-2

DO - 10.1007/s00034-014-9744-2

M3 - Article

AN - SCOPUS:84902828135

VL - 33

SP - 2251

EP - 2265

JO - Circuits, Systems, and Signal Processing

JF - Circuits, Systems, and Signal Processing

SN - 0278-081X

IS - 7

ER -