Robustness conditions of the LMS algorithm with time-variant matrix step-size.

DOI10.1016/S0165-1684(00)00088-8MaRDI QIDQ1575757zbMATH OpenFDO

Authors N. Delaunay, Markus Rupp, Jürgen Cezanne

Publication date 21 August 2000

Published in Signal Processing (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1016/S0165-1684(00)00088-8

zbMATH Keywords

Cellular phoneChannel estimationLeast-mean-squares algorithmTime-variant step-size matrix

Mathematics Subject Classification ID

Estimation and detection in stochastic control theory (93E10) Communication theory (94A05)

Review text

Gradient-type algorithms commonly employ a scalar step-size, i.e., each entry of the regression vector is multiplied by the same value before updating the coefficients. More flexibility, however, is obtained when this step-size is of matrix size. It allows not only for the individual scaling of the entries of the regression vector but rotations and decorrelations are possible as well due to the choice of the matrix. A well-known example for the use of a fixed step-size matrix is the Newton-LMS algorithm. For such a fixed step-size matrix, conditions are well known under which a gradient-type algorithm converges. This article, however, presents robustness and convergence conditions for a least-mean-square (LMS) algorithm with time-variant matrix step-size. An example of a channel estimator used in a cellular hand-phone is given that shows that the choice of a particular step-size matrix leads to considerable improvement over the fixed step-size case.

Recommendations

• Convergence of matrix LMS algorithms with applications to LMS/Newton
• New gradient-based variable step size LMS algorithms
• A simple variable step size LMS adaptive algorithm
• Extension of LMS stability condition over a wide set of signals
• Variable step-size normalized LMS algorithm by approximating correlation matrix of estimation error

Cited in

(7)

• The error variance of LMS with time-varying weights
• A New Robust Variable Step-Size NLMS Algorithm
• ROBUSTNESS ANALYSIS OF THE DATA-SELECTIVE VOLTERRA NLMS ALGORITHM
• Extension of LMS stability condition over a wide set of signals
• Convergence of matrix LMS algorithms with applications to LMS/Newton
• Robust continuous-time matrix estimation under dependent noise perturbations: sliding modes filtering and LSM with forgetting
• Variable step-size normalized LMS algorithm by approximating correlation matrix of estimation error

This page was built for publication: Robustness conditions of the LMS algorithm with time-variant matrix step-size.

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1575757)