Performance Analysis of Nonlinear-Predictive-Filter-Based Chaotic Synchronization

Abstract

Synchronization of chaotic systems is essential for chaotic communication schemes. If the communication environment is modeled stochastic, the extended-Kalman-filter-(EKF)-based synchronization scheme possesses minimum error in estimating the states for certain chaotic systems/maps. However, its intrinsic nonlinear approximation error can cause divergence and, hence, desynchronization. In this brief, a nonlinear predictive filter (NPF) is proposed for chaotic synchronization. The condition for stability and an approximate expression for the total normalized mean square error are derived. Numerical evaluations reveal that the proposed NPF-based synchronization scheme has better performance compared with the EKF-based scheme