A new approach to blind separation of cyclostationary sources

This paper studies the problem of blind source separation (BSS) under the assumption that the source signals are cyclostationary. Attention is restricted to methods based on second-order cyclostationary statistics (SOCS). Necessary and sufficient conditions for SOCS-based identifiability and SOCS-based separability are presented. Algorithms are developed to achieve BSS using only SOCS.

[1]  William A. Gardner,et al.  Spectral self-coherence restoral: a new approach to blind adaptive signal extraction using antenna arrays , 1990, Proc. IEEE.

[2]  Shun-ichi Amari,et al.  Why natural gradient? , 1998, Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP '98 (Cat. No.98CH36181).

[3]  Y.-C. Liang,et al.  Blind source separation using second-order cyclic-statistics , 1997 .

[4]  W. Gardner.  Exploitation of spectral redundancy in cyclostationary signals , 1991, IEEE Signal Processing Magazine.

[5]  William A. Gardner,et al.  Maximum likelihood and common factor analysis-based blind adaptive spatial filtering for cyclostationary signals , 1993, 1993 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[6]  Aníbal R. Figueiras-Vidal,et al.  An adaptive beamforming technique based on cyclostationary signal properties , 1995, IEEE Trans. Signal Process..

[7]  Jean-François Cardoso,et al.  Equivariant adaptive source separation , 1996, IEEE Trans. Signal Process..

[8]  Eric Moulines,et al.  A blind source separation technique using second-order statistics , 1997, IEEE Trans. Signal Process..