Analysis of a stabilization technique for the fixed-point prewindowed RLS algorithm

Author

Adali, Tulay ; Ardalan, Sasan H.

Author_Institution

Center for Commun. & Signal Process., North Carolina State Univ., Raleigh, NC, USA

Volume

39

Issue

9

fYear

1991

fDate

9/1/1991 12:00:00 AM

Firstpage

2079

Lastpage

2082

Abstract

A stable finite precision recursive least squares (RLS) algorithm is derived for the prewindowed growing memory case (forgetting factor, λ=1). The prewindowed growing memory RLS algorithm diverges under fixed-point implementation. The random walk phenomenon due to roundoff errors in the weight update causes the divergence of the algorithm. To overcome this effect, these roundoff errors are modeled such that their effect is incorporated into the algorithm. The steady-state behavior of this new algorithm is analyzed, and it is shown that the divergence phenomenon is actually eliminated, and the new algorithm converges

Keywords

convergence of numerical methods; least squares approximations; signal processing; convergence; divergence phenomenon; fixed-point implementation; forgetting factor; prewindowed RLS algorithm; prewindowed growing memory case; random walk phenomenon; recursive least squares; roundoff errors; signal processing; stabilization technique; stable finite precision RLS algorithm; steady-state behavior; weight update; Additive noise; Additive white noise; Algorithm design and analysis; Kalman filters; Least squares methods; Linear systems; Noise measurement; Resonance light scattering; Roundoff errors; Signal processing algorithms;

fLanguage

English

Journal_Title

Signal Processing, IEEE Transactions on

Publisher

ieee

ISSN

1053-587X

Type

jour

DOI

10.1109/78.134439

Filename

134439