A fast algorithm for blind separation of sources based on the cross-cumulant and Levenberg-Marquardt method

Author

Mansour, A. ; Barros, A. Kardec ; Kawamoto, M. ; Ohnishi, N.

Author_Institution

RIKEN, Bio-Mimetic Control Res. Center, Nagoya, Japan

fYear

1998

fDate

1998

Firstpage

323

Abstract

A new algorithm for the instantaneous mixture of the blind separation of sources problem is derived. This algorithm deals with multi-input multi-output (MIMO) channels. The cost function proposed in this paper can be considered as an extension of that previously proposed by us for only two sources and two sensors. The cost function is based on the cross-cumulant 2×2 and it is minimized using the Levenberg-Marquardt method. Generally, algorithm convergence is attained in less than 50 iterations and the experimental results are satisfactory even with five stationary sources and two non-stationary sources

Keywords

MIMO systems; convergence of numerical methods; higher order statistics; iterative methods; minimisation; signal processing; Levenberg-Marquardt method; MIMO channels; algorithm convergence; blind source separation; cost function; cross-cumulant; instantaneous mixture; iterations; minimization; multi-input multi-output channels; non-stationary sources; stationary sources; Biosensors; Cost function; MIMO; Minimization methods; Signal processing algorithms; Statistics;

fLanguage

English

Publisher

ieee

Conference_Titel

Signal Processing Proceedings, 1998. ICSP '98. 1998 Fourth International Conference on

Conference_Location

Beijing

Print_ISBN

0-7803-4325-5

Type

conf

DOI

10.1109/ICOSP.1998.770216

Filename

770216