ScholarLens Back to search
2010•Journal of tianjin University of TechnologyRequires access

Novel variable step size LMS algorithm

Hongwei Dai

Open publisher page 1 citations

Abstract

Some variable step size LMS algorithms in literature are analyzed,based on which an improved one is presented.This algorithm builds a nonlinear relationship between the step size,the error signal and the weighting coefficients,and then the asynchronous update of the weight vector becomes optimal.Simulation results show that this algorithm has higher speed of convergence,lower steady state error and more stable convergence process.

About this research paper

What this paper is about

Some variable step size LMS algorithms in literature are analyzed,based on which an improved one is presented.This algorithm builds a nonlinear relationship between the step size,the error signal and the weighting coefficients,and then the asynchronous update of the weight vector becomes optimal.Simulation results show that this algorithm has higher speed of convergence,lower steady state error and more stable convergence process.

Why it matters

OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Some variable step size LMS algorithms in literature are analyzed,based on which an improved one is presented.This algorithm builds a nonlinear relationship between the step size,the error signal and the weighting coefficients,and then the asynchronous update of the weight vector becomes optimal.Simulation results show that this algorithm has higher speed of convergence,lower steady state error and more stable convergence process.

Key concepts: Convergence (economics), Weighting, Algorithm, Variable (mathematics), Asynchronous communication, Nonlinear system, Computer science, Process (computing)

Related papers

Related paper 1Related paper 2Related paper 3
Back to paper searchBrowse research topicsOriginal source
Novel variable step size LMS algorithm — Research Paper | ScholarLens