DocumentCode :
703315
Title :
On the issue of rank estimation in subspace tracking: The NA-CSVD solution
Author :
Pango, Philippe A. ; Champagne, Benoit
Author_Institution :
INRS-Telecommun., Univ. du Quebec, Verdun, QC, Canada
fYear :
1998
fDate :
8-11 Sept. 1998
Firstpage :
1
Lastpage :
4
Abstract :
The issue of rank estimation in subspace tracking algorithms is adressed. In a recent paper, we proposed a subspace tracking algorithm, the NA-CSVD. We now extend the performance of NA-CSVD to rank tracking by coupling it with a recently proposed rank tracking technique. The paper includes an overview of typical rank+subspace tracking algorithms which, along with our proposed algorithm, are tested in various simulation scenarios. The new algorithm tracks efficiently the rank and the signal subspace.
Keywords :
estimation theory; singular value decomposition; target tracking; NA-CSVD; noise average cross-terms singular value decomposition; rank estimation; rank+subspace tracking algorithm; signal subspace; Approximation algorithms; Delays; Direction-of-arrival estimation; Estimation; Noise; Signal processing algorithms;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Signal Processing Conference (EUSIPCO 1998), 9th European
Conference_Location :
Rhodes
Print_ISBN :
978-960-7620-06-4
Type :
conf
Filename :
7089786
Link To Document :
بازگشت