Convergence behavior of affine projection algorithms
Authors:
- A.A.L. BeexABA.A.L. Beex
- 1 April 2000
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 48 (4) , 1086-1096
Abstract
A class of equivalent algorithms that accelerate the convergence of the normalized LMS (NLMS) algorithm, especially for colored inputs, has previously been discovered independently. The affine projection algorithm (APA) is the earliest and most popular algorithm in this class that inherits its name. The usual APA algorithms update weight estimates on the basis of multiple, unit delayed, input signal vectors. We analyze the convergence behavior of the generalized APA class of algorithms (allowing for arbitrary delay between input vectors) using a simple model for the input signal vectors. Conditions for convergence of the APA class are derived. It is shown that the convergence rate is exponential and that it improves as the number of input signal vectors used for adaptation is increased. However, the rate of improvement in performance (time-to-steady-state) diminishes as the number of input signal vectors increases. For a given convergence rate, APA algorithms are shown to exhibit less misadjustment (steady-state error) than NLMS. Simulation results are provided to corroborate the analytical resultsKeywords
This publication has 11 references indexed in Scilit:
- Normalized LMS algorithm with orthogonal correction factors2002
- The fast affine projection algorithm2002
- Fast generalized affine projection algorithmInternational Journal of Adaptive Control and Signal Processing, 2000
- A family of adaptive filter algorithms with decorrelating propertiesIEEE Transactions on Signal Processing, 1998
- Underdetermined-order recursive least-squares adaptive filtering: the concept and algorithmsIEEE Transactions on Signal Processing, 1997
- On a class of computationally efficient, rapidly converging, generalized NLMS algorithmsIEEE Signal Processing Letters, 1996
- On the convergence behavior of the LMS and the normalized LMS algorithmsIEEE Transactions on Signal Processing, 1993
- Variable length stochastic gradient algorithmIEEE Transactions on Signal Processing, 1991
- Convergence and performance analysis of the normalized LMS algorithm with uncorrelated Gaussian dataIEEE Transactions on Information Theory, 1988
- An adaptive filtering algorithm using an orthogonal projection to an affine subspace and its propertiesElectronics and Communications in Japan (Part I: Communications), 1984