On the performance of real-time traffic in a wireless environment

Abstract

To gain an understanding of the impact of the wireless channel on performance, we consider three different models. In the simplest model, the errors are assumed to be independent from slot t o slot. Next, a binary Markov channel is considered, where the channel is modeled via a two-state Markov chain, the error rate being a function of the state. This model would not allow for high variability in the error rate. To address this issue, a final ternary Markov channel model is considered. The arrival process is modeled as a superposition of a finite number of periodic streams. As a final complexity to the model, probabilistic arrivals for each stream are considered.