Computational aspects of adaptive radial basis function equalizer design

This paper investigates the computational aspects of radial basis function (RBF) equalizers. In an RBF implementation of the Bayesian equalizer the RBF centers are placed at equalizer channel states and the output layer weights are adjusted to +1/-1. Here we propose an RBF equalizer with scalar centers which can implement the Bayesian decision function. The proposed RBF equalizer provides lower computational complexity compared to the reported RBF equalizers and can efficiently employ subset center selection for computing the decision function resulting in a substantial reduction in computational complexity