Cascaded structures for blind source recovery

Blind Source Recovery (BSR) is an interesting autonomous and unsupervised stochastic adaptation problem that includes the well-known blind adaptive problems of Blind Source Separation (BSS), Deconvolution (BSD) and Equalization (BSE). BSR includes also the nonlinear case and hence focuses on reproducing or estimating the source signals even if environment identification is not achieved. A number of outstanding research contributions have been made in this field, however, the issues of application are still in their infancy. Most of the BSR algorithms have characteristics, which make them suitable for a particular subclass of problems. In order to develop a generalized source recovery framework and yet achieve optimal performance in all cases, there is a need to explore further architectural and/or algorithmic domains. In this paper, we approach this goal in the architecture domain by focusing on the use of cascaded structures for BSR. The paper discusses the need, choice, possible forms and properties of several cascaded structures. Some illustrative simulations have been included to highlight the advantages of some of the proposed structures.

[1]  Fathi M. A. Salam,et al.  Blind source recovery: algorithms for static and dynamic environments , 2001, IJCNN'01. International Joint Conference on Neural Networks. Proceedings (Cat. No.01CH37222).

[2]  Khurram Waheed,et al.  State Space Feedforward and Feedback Structures for Blind Source Recovery , 2001 .

[3]  Seungjin Choi,et al.  ON-LINE SEQUENTIAL MULTICHANNEL BLIND DECONVOLUTION: A DEFLATION APPROACH , 2002 .

[4]  Inbar Fijalkow,et al.  Blind adapted, pre-whitened constant modulus algorithm , 2001, ICC 2001. IEEE International Conference on Communications. Conference Record (Cat. No.01CH37240).

[5]  Khurram Waheed,et al.  Blind source recovery: some implementation and performance issues , 2001, Proceedings of the 44th IEEE 2001 Midwest Symposium on Circuits and Systems. MWSCAS 2001 (Cat. No.01CH37257).

[6]  Khurram Waheed,et al.  State space blind source recovery for mixtures of multiple source distributions , 2002, 2002 IEEE International Symposium on Circuits and Systems. Proceedings (Cat. No.02CH37353).

[7]  G. Erten,et al.  Formulation and algorithms for blind signal recovery , 1997, Proceedings of 40th Midwest Symposium on Circuits and Systems. Dedicated to the Memory of Professor Mac Van Valkenburg.