Sensitivity of wireless network simulations to a two-state Markov model channel approximation

In this paper, we investigate the appropriateness of wireless network simulations that employ a two-state Markov model to approximate a flat Rayleigh fading channel. Our approach first analyzes the statistical similarities and differences between block error processes generated by 1) an 802.11b defined communication scheme in the presence of flat Rayleigh fading and 2) a representative two-state Markov model. It is shown that, at low SNR, the Markov model does not generate an adequate frame error process. Furthermore, we quantify the effects of said statistical deviations through ns2 network simulations. This work further advances the understanding of the appropriate, and inappropriate, environments under which the two-state Markov model is a suitable approximation of a fading channel.