Date of Award

2009

Publication Type

Master Thesis

Degree Name

M.A.Sc.

Department

Electrical and Computer Engineering

Keywords

Applied sciences

Supervisor

Dr. Kwan

Rights

info:eu-repo/semantics/openAccess

Abstract

The system identification problem is notably dealt with using adaptive filtering approaches. In many applications the unknown system response consists of an initial sequence of zero-valued coefficients that precedes the "active" part of the response. The presence of these coefficients introduces a flat delay in the incoming signals which can take significantly large values. When most adaptive approaches attempt to model such a system, the presence of flat delay impairs their operation and performance. The approach introduced in this thesis aims to model the flat delay and "active" part of the unknown system separately. An efficient system for time delay estimation (TDE) is introduced to estimate the flat delay of an unknown system. The estimated delay is then compensated within the adaptive system thus allowing the latter to cover the active part ofthe unknown system. The proposed system is applied to the Acoustic Echo Cancellation (ABC) problem.

Share

COinS