A Bilinear Neural Adaptive Equalizer for Chaos-Based Communication Channel

Based on analysis of Volterra series filter, combing characteristic of single layer neural network, and extending bilinear filter's structure, a bilinear neural network equalizer is studied in this paper for chaos-based communication systems, and adaptive algorithm is deduced by the NLMS. Computer simulations show that BilinearNN equalizer can availably remove nonlinear's distortions and ISI, and better performance also gets obviously improvement no matter what linear channel or nonlinear channel