Hierarchical Markov model for burst error analysis in wireless communications

This paper presents a two-layer hierarchical Markov model for a study of bursty statistics in wireless channels. In particular, by separating random errors from burst errors, an error sequence can be divided into two independent parts, burst sections and random sections. Consequently, we come to a new model with an interconnection of two Markov sub-models, one burst sub-model and one random sub-model. The validity of the model is confirmed by providing the model parameters in a Rayleigh fading channel. A comparison between the presented model and the simplified Fritchman (1965) model shows that this model can achieve a more precise characterization of burst properties in wireless channels