Nonlinear interference cancellation using a radial basis function network

Conventional linear filtering techniques cannot suppress interference or noise in the same band as the signal without degrading the signal. However if the corrupting noise arises from a nonlinear low dimensional dynamical system, it is possible to model the noise as a deterministic process rather than a stochastic one. In this paper a combination of linear and nonlinear models are used to separate the linear signal from the nonlinear noise. The normalised gaussian radial basis function (RBF) network is used to model the nonlinear interference. Decimators have been implemented to reduce the computational cost of the RBF network and re-embed the filtered chaos.