Àá½Ã¸¸ ±â´Ù·Á ÁÖ¼¼¿ä. ·ÎµùÁßÀÔ´Ï´Ù.
KMID : 0366519970170010147
Annual Bulletin Seoul Health
1997 Volume.17 No. 1 p.147 ~ p.164
A Study on The Control-Theoretic Improvement of The Least Mean Squares Algorithm
Chung Dong-Keun

Abstract
In signal processing and recognition systems, the LMS(Least Mean Squares) algorithm has played a key role because of its simple and easy of computation. However, the algorithm has disadvantage of the slow convergence rate. The purpose of this study is to develope the improved LMS algorithm that has the fast convergence rate and less misadjustment through the Kwong¡¯s analysis method, by which the algorithm is analyzed as a control system rather than the conventional statistical approach. As a result of this study, the Compensated LMS algorithm(CLMS) was proposed and the property of convergence rate of it was reinforced with the less misadjustment.
KEYWORD
LMS algorithm, MLMS algorithm, convergence rate, Misadjustment
FullTexts / Linksout information
Listed journal information