Journal of Guangdong University of Technology ›› 2007, Vol. 24 ›› Issue (2): 103-106.

• Comprehensive Studies • Previous Articles     Next Articles

Blind Signal Separation Algorithm Based on Quasi-Newton

  

  1.  (Faculty of Applied Mathematics,Guangdong University of Technology,Guangzhou 510006,China)
  • Online:2007-07-02 Published:2007-07-02

HTML

PDF (PC)

2133

Abstract: This paper proposes a new blind signal separation algorithm which uses joint diagonalization of correlation matrices as the cost function,and separates source signal based on quasi-Newton’s DFP method.The algorithm improves the convergence rate.Validity and performance of the proposed algorithm are demonstrated by extensive computer simulations.

Key words: blind signal separation； quasi-Newton method； joint diagonalization；

[1] Yeredor A.Non-orthogonal joint diagonalization in theleast-squares sense with application in blind source separa-tion. IEEE Transactions on Signal Processing . 2002

[2] Parra L,Spence C.Convolutive blind separation of non-sta-tionary sources. IEEE Trans.Speech and Audio Pro-cessing . 2000

[3] Tong L,Liu R W,Soon V C,et al.Indeterminacy andidentifiability of blind identification. IEEE Trans Cir-cuits and Systems . 1991

[4] Cardoso J F,Souloumiac A.Blind beamforming for non-Gaussian signals. IEE Proceedings F Radar and Signal Processing . 1993

[5] Belouchrani A,Meraim K A,Cardoso J F,et al.A blindsource separation technique using second-order statistics. IEEE Transactions on Signal Processing . 1997
No related articles found!
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
Copyright © 2017 Journal of Guangdong University of Technology
Add：729 East Dongfeng RD., Guangzhou PRC. Postcode: 510090
Tel：020-37626653,020-37626139,020-37626396
Email：xbzrb@gdut.edu.cn
Supported:Beijing Magtech