State estimation of chaotic stochastic systems with applications to chaotic communication

Huawei Ruan, Marquette University

Abstract

This work presents novel techniques for state estimation of nonlinear stochastic systems, especially chaotic stochastic systems, as well as their applications to chaotic communication systems. Several nonlinear estimation algorithms are developed, such as the Current Output Filter (COF), Unscented Current Output Filter (UCOF), Current Output Particle Filter (COPF) and Minimum Phase Space Distance (MPSD) for the purpose of state and parameter estimation for different chaotic stochastic systems. Their estimation performances are analyzed and compared with the classical nonlinear estimation algorithm, the Extended Kalman Filter. A new one-dimensional chaotic dynamical system, namely the generalized Tent map, is also proposed. This map has a very broad parameter region for generating uncorrelated chaotic sequences. The property of this map is used in this work to develop an M-ary Chaotic Shift Keying (MCSK) modulation scheme that can modulate a multi-bit digital symbol with just one chaotic sequence and chaotic CDMA system in which multi-user information is transmitted through the same channel. This MCSK modulation scheme greatly improves the information carrying capacity of communication system. The estimation techniques proposed in this work are employed for estimating the parameters of the chaotic sequences to demodulate the chaotically modulated digital symbols. The high orthogonality between the sequences generated from the generalized Tent map with different parameters is used to provide an effective new alternative approach for spreading code generation in CDMA systems. The effects of non-ideal channel characteristics in chaotic communication systems are also investigated in this work. In this case, the proposed nonlinear estimation techniques are shown to effectively identify the non-ideal channel model in chaotic communication systems.

This paper has been withdrawn.