Loughborough University
Leicestershire, UK
LE11 3TU
+44 (0)1509 263171
Loughborough University

Loughborough University Institutional Repository

Please use this identifier to cite or link to this item: https://dspace.lboro.ac.uk/2134/5613

Title: A new gradient based variable step-size LMS algorithm
Authors: Zhang, Yonggang
Li, Ning
Chambers, Jonathon
Issue Date: 2006
Publisher: © IEEE
Citation: ZHANG, Y., LI, N. and CHAMBERS, J.A., 2006. A new gradient based variable step-size LMS algorithm. IN: 2006 8th International Conference on Signal Processing (ICSP 2006), Guilin, China, Vol. 1, 16-20 November.
Abstract: A new gradient-based variable step-size LMS algorithm (VSSLMS) is proposed in this paper. The step size of the proposed algorithm is proportional to the squared norm of the smoothed gradient vector, which is a proximity measure of the adaptive process. In comparison with the existing methods, the proposed VSSLMS algorithm has improved convergence properties. Furthermore, the parameter choice of the proposed algorithm can be easily determined according to the analysis. Simulation results show the good properties of the proposed algorithm and support the performance analysis.
Description: This is a conference paper [© IEEE]. It is also available at: http://ieeexplore.ieee.org/ Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.
Version: Published
DOI: 10.1109/ICOSP.2006.344518
URI: https://dspace.lboro.ac.uk/2134/5613
Appears in Collections:Conference Papers and Contributions (Electronic, Electrical and Systems Engineering)

Files associated with this item:

File Description SizeFormat
A NEW GRADIENT BASED VARIABLE STEP-SIZE LMS ALGORITHM.pdf263.77 kBAdobe PDFView/Open

 

SFX Query

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.