DocumentCode :
547582
Title :
Modified sequential regression algorithm
Author :
Mehranpour, Mehdi ; Nourinia, Javad ; Ghobadi, Changiz
Author_Institution :
Dep. Of Electrical Engineering, University of Urmia, Urmia, Iran
fYear :
2011
fDate :
17-19 May 2011
Firstpage :
1
Lastpage :
4
Abstract :
In this paper, we propose an algorithm with higher speed at the start and lower misadjustment near the optimum point. The proposed algorithm consists of a matrix that multiplies in sequential regression (SER) algorithm factor and uses a new variable step size that provides a faster convergence for searching weights and output error to optimum point. As a result, it behaves like variable step size LMS (VSLMS) algorithm in terms of convergence speed.
Keywords :
Adaptive filters; Algorithm design and analysis; Approximation algorithms; Convergence; Equations; Least squares approximation; Signal processing algorithms; Misadjustment; adaptive filters; optimum point; sequential regression (SER); variable step size; variable step size LMS (VSLMS);
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Electrical Engineering (ICEE), 2011 19th Iranian Conference on
Conference_Location :
Tehran, Iran
Print_ISBN :
978-1-4577-0730-8
Electronic_ISBN :
978-964-463-428-4
Type :
conf
Filename :
5955470
Link To Document :
بازگشت