Recursive Least Absolute Error Algorithm: Analysis and Simulations

Shin'ichi KOIKE

IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences   Vol.E85-A    No.12    pp.2886-2893
Publication Date: 2002/12/01
Online ISSN: 
Print ISSN: 0916-8508
Type of Manuscript: PAPER
Category: Digital Signal Processing
RLS algorithm,  sign algorithm,  transient analysis,  steady-state performance,  impulse noise,  

Full Text: PDF>>
Buy this Article

Recursive least absolute(RLA) error algorithm is derived which is basically the sign algorithm (SA) combined with recursive estimation of the inverse covariance matrix of the reference input. The name RLA comes from the absolute error criterion. Analysis of the transient behavior and steady-state performance of the RLA algorithm is fully developed. Results of experiment show that the RLA algorithm considerably improves the convergence rate of the SA while preserving the robustness against impulse noise. Good agreement between the simulation and the theoretically calculated convergence validates the analysis.