Non-linear channel equalisation using minimal radial basis function neural networks

Abstract

This paper presents the study results of non-linear channel equalisation problems in data communications using a minimal radial basis function neural network structure, referred to as MRAN (minimal resource allocation network). The MRAN algorithm uses on-line learning and has the capability to grow and prune the RBF network's hidden neurons ensuring a parsimonious network structure. Compared to earlier methods, the proposed scheme does not have to estimate the channel order first, and fix the model parameters. Results showing the superior performance of the MRAN algorithm for two different non-linear channel equalisation problems, along with a linear non-minimum phase problem, are presented.