Dissertation/Thesis Abstract

Entrainment: Adaptive feedback cancelling electro-acoustic system dynamics
by Theverapperuma, Lalin Shrimal, Ph.D., University of Minnesota, 2007, 165; 3268352
Abstract (Summary)

Adaptive feedback cancellation is a common method of cancelling acoustic feedback in a variety of electro-acoustic systems. The time varying acoustic feedback leakage paths can only be eliminated with adaptive feedback cancellation. When an electro-acoustic system with an adaptive feedback canceller is presented with a correlated input signal, a recurrent distortion artifact ( entrainment) is generated.

By using nonlinear analysis and iterating the equations derived for a simplified feedback canceller, the entrainment phenomenon is explained, and the inherency of it to feedback cancellation configurations is confirmed. Further clarifications are given using the mean square error (MSE) surface concepts. The main application of this research is to avoid feedback problems in public announcing systems and personal electro-acoustic systems.

The problem of entrainment known for decades as coloration artifact, bias, and low-frequency artifact in the electro-acoustic industry, has been attributed to a correlated input signal, although the cause had not been identified. One section of this thesis unifies previous approaches to solving the entrainment problem and discusses their limited effectiveness. Finally, four novel non entraining feedback canceller algorithms have been developed based on the theoretical findings of this dissertation.

Indexing (document details)
Advisor: Holte, James E.
Commitee: Georgiou, Tryphon, Parhi, Keshab
School: University of Minnesota
Department: Electrical Engineering
School Location: United States -- Minnesota
Source: DAI-B 68/05, Dissertation Abstracts International
Source Type: DISSERTATION
Subjects: Electrical engineering
Keywords: Adaptive feedback cancelling, Electroacoustic, Entrainment, Feedback cancellation
Publication Number: 3268352
ISBN: 9780549070016
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest