Utilization of finite-alphabet (FA) signals for the underdetermined blind source separation in a weakly nonlinear channel

In this paper, an underdetermined blind source separation in a weakly static (or memoryless) nonlinear channel with multi-input and single-output (MISO) is proposed. For that purpose, the finite alphabet coding, and a real-time self-organizing map algorithm are utilized. Finally, some simulation results, obtained by considering application of the proposed approach to power-line communications, are provided to demonstrate its validity.