Extension of joint diagonalization for moving targets

Author

Hori, Gen ; Tanaka, Toshihisa

Author_Institution

Fac. of Bus. Adm., Asia Univ., Musashino, Japan

fYear

2009

fDate

13-16 Dec. 2009

Firstpage

309

Lastpage

311

Abstract

Joint diagonalization algorithm is extended for a time-varying mixing matrix. A new method termed Â¿differential joint diagonalization (DJD)Â¿ is introduced to calculate the time-difference of the demixing matrix from the time-difference of the time-delayed correlations and applied to BSS algorithm based on joint diagonalization of time-delayed correlations. An alternate algorithm of differential joint diagonalization (DJD) and joint diagonalization (JD) for tracking a time-varying mixing matrix is proposed.

Keywords

blind source separation; correlation methods; matrix algebra; BSS algorithm; blind source separation algorithm; differential joint diagonalization method; joint diagonalization algorithm; time difference method; time-delayed correlations; time-varying mixing matrix; time-varying mixing matrix tracking; Adaptive signal processing; Agriculture; Approximation algorithms; Asia; Blind source separation; Conferences; Delay effects; Linear approximation; Matrices; Signal processing algorithms;

fLanguage

English

Publisher

ieee

Conference_Titel

Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP), 2009 3rd IEEE International Workshop on

Conference_Location

Aruba, Dutch Antilles

Print_ISBN

978-1-4244-5179-1

Electronic_ISBN

978-1-4244-5180-7

Type

conf

DOI

10.1109/CAMSAP.2009.5413270

Filename

5413270