A Novel UPF-Based Demodulator for Chaotic Parameter Modulation Systems

Chaotic parameter modulation is one kind of secure communications in which the system noise is merged into the dynamic evolution of modulating signal and the observed sequence is a chaotic signal plus non-Gaussian noise. So far the known demodulation algorithms for the sequence are limited to the Gaussian noises and high signal to noise ratios are required. To solve these issues, the idea of PF (particle filter) is introduced to deal with these issues in which the distribution of the noise could be any form. The state equations in UPF (unscented particle filter) form for the demodulation of chaotic parameter modulations are derived, and a novel modified algorithm based on UPF for the demodulation is presented. Because of the utilization of the nonlinear innovation hidden in the received chaotic signal and non-Gaussian noise, there is no limit to the Gaussian noise or not, and the requirement of signal to noise ratio is obviously decreased in the same time. The simulated results show that the presented algorithm can demodulate the messages from the received sequence accurately under the relatively low signal to noise ratio. The demodulation comparison with PF (particle filter) is also carried out and results show that the MSE(mean square error) of the UPF-based algorithm is better than PF about 10 dB.