Analytical Modeling And Performance Evaluation Of IEEE 802.11 Based MAC Protocol In Wireless Sensor Networks

With the rapid development of Information Technology, the wireless network communication technology embraces its fast development and becomes more and more popular these years, as the supplement and expansion form of wired networks. Many new technologies represented by wireless local area networks (WLAN) have emerged in this field. And most recently, the wireless sensor networks (WSN), a network comprised of many mobile devices with sensing function and used for collecting, processing and transmitting information, has be generated.Within those wireless networks, performance evaluation of the media access control (MAC) Layer plays a very important role for many design issues such as scheduling, admission control, and resource allocation for quality of service (QoS) provisioning, as well as serves as the theoretical foundation for parameter design and protocol improvement. Hence the analytical modeling of MAC protocols always remains the hot-spot in the research of wireless networks. In addition, with it specific feature like energy limitation and large amount of nodes, the analysis of MAC protocol in wireless sensor networks confronts many new problems and challenges, which is quite different from those in traditional wired wireless networks.In this article, we focus on analyzing the MAC protocol in wireless networks and present two new analytical models; one for IEEE 802.11 MAC protocol and the other for S-MAC protocol. By using our model, we further evaluate their performance.In our analytical model for IEEE 802.11 MAC, we mainly study the distributed coordination function (DCF) in IEEE 802.11 MAC, because the other function point coordination function (PCF) is an optional function and the share of the channel in this function is free of contentions. By noticing that there is an inconsistence between the commonly used assumption in most of the exist models and the specific requirement in the protocol regulation, we proposed a new event model based on a Markov chain to study the IEEE 802.11 MAC protocol and solve the problem caused by the inconsistence which results in that the occurrences of events on the channel are not independent of each other. Simulation results demonstrate that our proposed model is more accurate than Bianchi's new model-which was presented to tackle the"inconsistence"problem in 2005-in evaluating the throughput and the average packet delay. Moreover, our model can further derive the distribution of packet delay and the energy efficiency, thus enabling us to have a more thorough understanding of the protocol performance. In our analytical model for S-MAC, we analyze the mechanism designed for saving the energy cost by collision, over-hearing and idle-listening. Due to the periodic feature in S-MAC protocol, we can not simply build the models similar to those used in IEEE 802.11. So, by considering the periodic feature and the independence between the behaviors in different cycles which results from the backoff cancellation, our model studies the impact of several factors together on the protocol performance, including periodical listen and sleep, different traffic load, backoff mechanism, queuing behavior at the MAC layer, imperfect channel condition and the dependence of non-independent nature of service delay distribution of nodes. The proposed model for S-MAC consists of two sub-models: one is a Markov chain model to study the node behavior, and the other is M/G/1/∞model to study the queuing behavior. Using our model, we can compute two important probabilities, which are the successful transmission probability of a node in a cycle and the probability that a successfully transmitting node has a backoff counter initially set to i. Based on the obtained conclusion, we can analyze the throughput,service delay, energy consumption and energy efficiency. Furthermore, we can study the impact of the two important protocol parameters (contention window size and duty cycle) on the trade off between QoS and energy consumption. Finally, the accuracy of our model is validated by NS-2 simulator.