A non-stationary hidden Markov model for satellite propagation channel modeling

The hidden Markov models are widely used for modeling random processes. Drawbacks in a standard hidden Markov model for modeling state occupancy have often been pointed out. In this paper, we proposed a propagation channel model for predicting the fading properties of satellite-to-Earth propagation channels using non-stationary hidden Markov model approach. An accurate prediction model was developed to provide better performance in modeling channel properties than a standard hidden Markov model and validated by several measured fading signal series. A close match between simulated and measured data shows that the proposed model predicts satisfactorily the first and second order statistics of propagation properties in several user environments. Considering the state stay duration issues, our proposed method can produce more accurate simulation results and extend efficiently the applicability of the HMM approach in modeling propagation channel properties for land mobile satellite communication services.