A particle swarm optimization-least mean squares algorithm for adaptive filtering

Author

Krusienski, D.J. ; Jenkins, W.K.

Author_Institution

Dept. of Electr. Eng., Pennsylvania State Univ., University Park, PA, USA

Volume

1

fYear

2004

fDate

7-10 Nov. 2004

Firstpage

241

Abstract

A particle swarm optimization-least mean squares (PSO-LMS) algorithm is presented for adapting various classes of filter structures. The LMS algorithm is widely accepted as the preeminent adaptive filtering algorithm because of its speed, efficiency and provably convergent local search capabilities. However, for multimodal error surfaces, a global search algorithm, such as PSO or the genetic algorithm (GA), is required. The proposed PSO-LMS hybrid algorithm combines the advantageous properties of the two conventional algorithms to provide enhanced performance characteristics.

Keywords

adaptive filters; algorithm theory; least mean squares methods; optimisation; stochastic processes; LMS; PSO; adaptive filtering algorithm; global search algorithm; least mean square algorithm; multimodal error surface; particle swarm optimization; Adaptive filters; Filtering algorithms; Genetic algorithms; IIR filters; Least squares approximation; Neural networks; Particle swarm optimization; Polynomials; Signal processing algorithms; Stochastic processes;

fLanguage

English

Publisher

ieee

Conference_Titel

Signals, Systems and Computers, 2004. Conference Record of the Thirty-Eighth Asilomar Conference on

Print_ISBN

0-7803-8622-1

Type

conf

DOI

10.1109/ACSSC.2004.1399128

Filename

1399128