On the tracking performance of certain algorithms based on a time-varying nonlinear Hammerstein model

An Extended Kalman Filter, EKF and a Recursive Prediction Error Method, RPEM based on a time-varying nonlinear Hammerstein model are proposed. These algorithms can be applied to the problem of acoustic echo cancelation. The performance of the suggested methods is compared to time-varying Cramer-Rao lower bounds, CRB's which are also derived.