Delay-Transmission Probability Tradeoffs in IEEE 802.11 based Ad hoc Networks

Several factors contribute towards the delay performance in a distributed ad hoc network. One such factor is the probability at which individual nodes attempt to access the shared medium. Moreover, the lack of coordination among the nodes make it even more difficult to predict the delay performance of the system. A node's decision whether to transmit a packet or not, determines not only the expected performance of the node but also the network as a whole. In this paper, we consider an IEEE 802.11 based ad hoc network where the nodes employ a p-persistent model in their MAC protocol. First, we present an analysis for the delay and determine the probability mass function of the delay between any two consecutive successful transmissions. We then show that there is an optimal transmission probability that each node chooses in order to maximize the probability for achieving a target delay. Through numerical results, we compare the p- persistent model with slotted aloha model to show that despite different models, the delay-transmission probability tradeoffs have similar characteristics. An easy way to calculate the near optimal transmission probability is proposed. We further discuss ways to control delay using transmission probability in IEEE 802.11 based ad hoc networks.