An analytical model for evaluating usable throughput in ad-hoc wireless networks

Abstract

This paper develops an analytical model for evaluating usable throughput in wireless ad-hoc networks. Our analysis approach uses a Markov chain model to describe the wireless channel. Each state represents the number of simultaneous transmissions in the network. The advantage of this model is that it allows us determine the probability of being in each of the channel states. This model was previously used for evaluating the Quality of Service and Signal to Interference Ratio. We start by re-deriving the per user throughput capacity, first proposed by Gupta and Kumar, using our system model. We then use our model to derive the usable throughput. We define the usable throughput as the maximum throughput in a network with a high packet completion rate. These two results are compared to the per user throughput capacity using numerical methods. As expected, the curve for the usable throughput lies below the one for the maximum throughput.