2005Systems engineering and electronicsRequires access

New variable step-size LMS adaptive algorithm

Tang Pu-ying

Open publisher page 9 citations

Abstract

This paper discusses about a type of algorithms of variable step-size LMS adaptive filtering and presents another new non-linear functional relationship between the step-size and the error signal,which improves the algorithm performance.On the condition of the same convergence properties or the same excess MSE, the new algorithm has less excess MSE or has better convergence properties than the former algorithms.Computer simulation results confirm the theoretical analysis and show the algorithm is superior to the former algorithms in performance.

About this research paper

What this paper is about

This paper discusses about a type of algorithms of variable step-size LMS adaptive filtering and presents another new non-linear functional relationship between the step-size and the error signal,which improves the algorithm performance.On the condition of the same convergence properties or the same excess MSE, the new algorithm has less excess MSE or has better convergence properties than the former algorithms.Computer simulation results confirm the theoretical analysis and show the algorithm is superior to the former algorithms in performance.

Why it matters

OpenAlex reports 9 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

This paper discusses about a type of algorithms of variable step-size LMS adaptive filtering and presents another new non-linear functional relationship between the step-size and the error signal,which improves the algorithm performance.On the condition of the same convergence properties or the same excess MSE, the new algorithm has less excess MSE or has better convergence properties than the former algorithms.Computer simulation results confirm the theoretical analysis and show the algorithm is superior to the former algorithms in performance.

Key concepts: Algorithm, Convergence (economics), Adaptive filter, Variable (mathematics), Least mean squares filter, Adaptive algorithm, Computer science, Rate of convergence

Related papers

Back to paper searchBrowse research topicsOriginal source
New variable step-size LMS adaptive algorithm — Research Paper | ScholarLens