Adaptive RLS lattice filters for fast nonstationary signals

Author

Settineri, R. ; Favier, G.

Author_Institution

Nice Univ., France

fYear

1990

fDate

3-6 Apr 1990

Firstpage

1807

Abstract

An adaptive recursive-least-squares (RLS) lattice filter that is based on the adaptive normalized sliding-window covariance (ANSWC) algorithm is presented. The equations of this algorithm are derived by use of the projection approach. Two parameter change detectors are also presented. A Monte Carlo analysis of the ANSWC algorithm is carried out, and its performance is compared to that of the NSWC algorithm in terms of noise sensitivity and parameter tracking capability. The performance improvement obtained by using the ANSWC algorithm is shown in terms of the tradeoff between noise sensitivity and parameter tracking capability

Keywords

Monte Carlo methods; adaptive filters; filtering and prediction theory; least squares approximations; random noise; Monte Carlo analysis; adaptive RLS lattice filters; adaptive normalized sliding-window covariance; fast nonstationary signals; noise sensitivity; parameter change detectors; parameter tracking capability; projection approach; recursive-least-squares; Adaptive filters; Algorithm design and analysis; Change detection algorithms; Detectors; Equations; Lattices; Least squares methods; Monte Carlo methods; Performance analysis; Resonance light scattering; Signal processing algorithms;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1990. ICASSP-90., 1990 International Conference on

Conference_Location

Albuquerque, NM

ISSN

1520-6149

Type

conf

DOI

10.1109/ICASSP.1990.115842

Filename

115842