A blind source separation method based on Kalman filtering

Author

Hu, Zhihui ; Feng, Jiuchao

Author_Institution

Sch. of Electron. & Inf. Eng., South China Univ. of Technol., Guangzhou, China

fYear

2009

fDate

23-25 July 2009

Firstpage

473

Lastpage

476

Abstract

According to Nonlinear Principal Component Analysis (NPCA) criterion, a blind source separation algorithm based on Kalman filtering is proposed in this paper. The convergence property of the algorithm is analyzed. The performance of the algorithm is evaluated by using several different kinds of sources. The effect of the number of iteration steps and the observation noise for the performance are investigated. The results show that this algorithm can separate chaotic as well as other sources from linear instantaneous mixtures effectively.

Keywords

Kalman filters; blind source separation; convergence of numerical methods; iterative methods; principal component analysis; Kalman filtering; blind source separation algorithm; convergence property; iteration method; nonlinear principal component analysis; observation noise; Blind source separation; Chaotic communication; Filtering algorithms; Information filtering; Information filters; Kalman filters; Principal component analysis; Signal processing algorithms; Source separation; Vectors;

fLanguage

English

Publisher

ieee

Conference_Titel

Communications, Circuits and Systems, 2009. ICCCAS 2009. International Conference on

Conference_Location

Milpitas, CA

Print_ISBN

978-1-4244-4886-9

Electronic_ISBN

978-1-4244-4888-3

Type

conf

DOI

10.1109/ICCCAS.2009.5250473

Filename

5250473