Skip to Main Content
In this paper, we propose an expectation maximization (EM) based approach for semiblind identification of linear moving average (MA) systems. The system is driven by chaotic signals and a robust EM based estimator is formulated to estimate the system parameters and the driving chaotic signal. It is shown through numerical simulations that the proposed EM semiblind estimation technique outperforms other conventional techniques such as minimum nonlinear prediction error (MNPE) method and that based on extended Kalman filter (EKF). Also the proposed estimator is applied in the equalization of chaos based communication to illustrate the performance improvement.