A Novel Demodulation Method for Chaotic Parameter Modulation

This paper proposes a particle filter to demodulate the chaotic parameter modulation (CPM) signal CPM embeds the signal of transmission in the bifurcating parameter of a chaotic dynamical system, and uses the wide-band output of the chaotic system as the transmitted signal. As the bifurcating parameter of the chaotic dynamical system is unknown and the demodulation problem is just a system identification problem. Conventional methods like extended Kalman filter (EKF) and least mean square (LMS) have relatively poor performance. Incorporating Rao-Blackwellisaion (RB) and roughening noise method, particle filtering can be used to estimate the parameter sequentially. Simulation results confirm that the proposed particle filter has the improved performance compared to the EKF and LMS