Simplified Newton-type adaptive estimation algorithms

Author

Mavridis, Panagiotis P. ; Moustakides, George V.

Author_Institution

Dept. of Comput. Eng. & Inf., Patras Univ., Greece

Volume

44

Issue

8

fYear

1996

fDate

8/1/1996 12:00:00 AM

Firstpage

1932

Lastpage

1940

Abstract

A new adaptive estimation algorithm is presented. It is the result of a combination of the LMS and the fast Newton transversal filters (FNTF) class. The main characteristic of the proposed algorithm is its improved convergence rate as compared to LMS, for cases where it is known that LMS behaves poorly. This improved characteristic is achieved in expense of a slight increase in the computational complexity while the overall algorithmic structure is very simple (LMS type). The proposed algorithm seems also to compare relatively well against RLS and FNTF

Keywords

Newton method; adaptive estimation; adaptive filters; computational complexity; convergence of numerical methods; digital filters; filtering theory; least mean squares methods; LMS; Newton-type adaptive estimation algorithms; algorithmic structure; computational complexity; convergence rate; fast Newton transversal filters; Adaptive estimation; Computational complexity; Convergence; Echo cancellers; Helium; Least squares approximation; Predictive models; Resonance light scattering; Robustness; Transversal filters;

fLanguage

English

Journal_Title

Signal Processing, IEEE Transactions on

Publisher

ieee

ISSN

1053-587X

Type

jour

DOI

10.1109/78.533714

Filename

533714