Date of Award

Spring 1994

Document Type

Thesis

Degree Name

Master of Science in Electrical Engineering - (M.S.)

Department

Electrical and Computer Engineering

First Advisor

Yeheskel Bar-Ness

Second Advisor

Alexander Haimovich

Third Advisor

Zoran Siveski

Abstract

To increase the rate of convergence of the blind, adaptive, decision feedback equalizer based on the decorrelation criterion, structures have been proposed which dramatically increase the complexity of the equalizer. The complexity of an algorithm has a direct bearing on the cost of implementing the algorithm in either hardware or software. In this thesis, more computationally efficient structures, based on the fast transversal filter and lattice algorithms, are proposed for the decorrelation algorithm which maintain the high rate of convergence of the more complex algorithms. Furthermore, the performance of the decorrelation algorithm in a finite-precision environment will be studied and compared to the widely used LMS algorithm.

Share

COinS