Extended Kalman Filter-Based Codec for Chaotic Communication Systems
Abstract
Two methods are presented to encode and decode messages accurately with an expanded matrix representation of linear multivariable systems in which parameters are used to describe three-port chaotic secure communication networks. The matrix representation includes an optimal extended Kalman filter (EKF)-based observer, which is linear with time. The optimal linearization technique is used to find the exact linear models of the chaotic system at operating states of interest. Subsequently, the EKF algorithm is used to estimate the parameters and states in which a message is embedded. Using the EKF with the optimal linear model, the message can be adequately recovered at the receiver. However, the bit error rate is insufficiently small; therefore, presynchronizations and the carrier-digitalized sources are used to reduce the error rate to obtain robust communications. Numerical examples and simulation results demonstrate the effectiveness of the proposed methodology.
