Accurate residual wireless channel simulation using a two-tier Markov model

Accurate models of wireless channels are a necessary pre-requisite for correct simulation and analysis of wireless protocols, applications and systems. However, commonly used wireless channel models do not correctly capture the characteristics of a real channel mainly because these models ignore either the frame-loss process or the bit-error process. In this paper, we propose a novel two-tier Markov model in which a frame-level model is executed first and the bit-level model is excited only when the frame-level model produces an error. We compare the accuracy of the proposed model with several existing models using four independently collected sets of wireless (802.11 and 802.15.4) residual bit-error traces. We validate the accuracy of the proposed model using the bit-error rate, frame-error rate, cumulative distributions of error and error-free bursts, and information divergence measures. We show that the proposed two-tier Markov model generally provides better accuracy than other models.