Floating-Point Scaling for BSS Gain Control

In Blind Source Separation (BSS) processing, taking advantage of scaling factor indetermination and based on the floatingpoint representation, we propose a scaling technique applied to the separation matrix, to avoid the saturation or the weakness in the recovered source signals. This technique performs an Automatic Gain Control (AGC) in an on-line BSS environment. We demonstrate the effectiveness of this technique by using the implementation of a division free BSS algorithm with two input, two output. This technique is computationally cheaper and efficient for a hardware implementation.

Authors:

• Abdelmalek Fermas
• Adel Belouchrani
• Otmane Ait Mohamed

Keywords:

• Automatic Gain Control
• Blind Source Separation
• Floating-Point Representation
• FPGA Implementation.

References:

[1] B. Fagin and C. Renard, "Field Programmable Gate Arrays and Floating
Point Arithmetic," in IEEE Trans. on VLSI Systems, Vol. 2, Sept. 1994,
pp. 365-367.
[2] J.-P. Deschamps and G. Sutter, "Finite Field Division Implementation,"
in 15th Int. Conf. on Field Programmable Logic and Applications, 2005.
[3] A. Belouchrani, E. Bourennane and K. Abed-Meraim, "A closed form
Solution for the Blind Separation of Two Sources from Two Sensors," in
14th European Signal Processing Conf., 2006.
[4] J. Detrey and F. de Dinechin, "A VHDL Library of LNS Operators," in
37th Asilomar Conf. on Signals, Systems and Computers, 2003.
[5] A. Fermas, A. Belouchrani and O. Aitmohmed, "Hardware Implementation
of Free Division Block-based BSS Algorithm," in 7th IEEE Int.
NEWCAS Conf., 2009.