A functional link artificial neural network for adaptive channel equalization

摘要: Abstract Application of artificial neural network (ANN) structures to the problem channel equalization in a digital communication system has been considered this paper. The difficulties associated with nonlinearities can be overcome by equalizers employing ANN. Because nonlinear processing signals an ANN, it is capable producing arbitrarily complex decision regions. For reason, ANN utilized for problem. A scheme based on functional link (FLANN) proposed task. performance along two other compared conventional LMS equalizer. Effect eigenvalue ratio input correlation matrix studied. From simulation results, observed that FLANN equalizer outperforms terms bit-error rate (BER) and attainable MSB level over wide range spread, signal noise nonlinearities.