Adaptive blind equalization for chaotic communication systems using extended-Kalman filter

In this paper, a blind-channel equalization technique for chaotic communications based on extended Kalman filtering (EKF) is proposed. Assuming that the channel coefficients of a fading and multipath channel can be described by an autoregressive (AR) model, blind-channel equalization is formulated as a mixed nonlinear parameter and state estimation problem. Nonlinear filters such as EKF can then be used to estimate the state of the mixed system which represents the original signal without channel distortion. The stability problem of the proposed EKF equalization technique is also addressed. Simulations show that the proposed EKF equalization method has improved performance in terms of design, convergence speed, and demodulation quality compared to the standard synchronization-based technique for chaotic communications