ON ACCELERATING THE CONVERGENCE OF THE LMS ALGORITHM.

David W. Lin*

*Corresponding author for this work

Research output: Contribution to journalConference articlepeer-review

Abstract

The author studied a method of accelerating the LMS (least mean squares) algorithm in the context of echo cancellation and modified it to deal with the case in which the echo path loss is unknown beforehand. Several ways to remedy the stability problem resulting from an inaccurate estimation of the optimal gain are studied. The result is a switched-gain approach in which the adaptation gain switches between zero and a value determined from some signal statistics. The author also discusses the implementation of input whitening. Some topics for further research are suggested.

Original languageEnglish
Pages (from-to)171-175
Number of pages5
JournalConference Record - International Conference on Communications
StatePublished - 1 Dec 1986

Fingerprint Dive into the research topics of 'ON ACCELERATING THE CONVERGENCE OF THE LMS ALGORITHM.'. Together they form a unique fingerprint.

Cite this