Robust chaotic parameter modulation based on hybrid extended Kalman filter and hidden Markov model detector

An optimum demodulator for chaotic parameter modulation (CPM) is presented in this paper. A binary mapping model is used for time-varying bifurcation parameter of logistic map to increase the security of communication. Due to dynamic behavior of such a system, a combined extended Kalman filter (EKF) and hidden Markov model (HMM) is used as detector which provide time domain processing. EKF is used as a state vector estimator and HMM is used to assign the probability of each map and identify the most likely map to the active map. It will be shown that this approach provides up to 30 dB improvement on signal to noise ratio. This method reduced the complexity of EKF estimator and it is also independent of bifurcation parameter values. It is shown that proposed method is robust for various types of noise and bifurcation parameter values.