作者: Chin-Tsu Yen , Rui-Chang Lin , Wan-De Weng
DOI:
关键词:
摘要: A new improved soft-feedback functional link artificial neural-network (ISFFLANN) based nonlinear channel equalizer is proposed in this paper. By using the functional expansion utilities, the ISF-FLANN does not need the hidden layers, which are existed in most of the multilayer perceptron network (MLP)-based equalizers. So the ISF-FLANN exhibits much simpler structure and thus requires less amount of computation during the training mode. We find that the use of soft feedback can greatly improve the performance of our previous work …