Data Traffic Modeling of ML-MAC for Wireless Sensor Networks

Abstract

Wireless sensors collect the data and transmit it to the base station through a network of nodes. The flow of data through the network can also be termed as traffic flow. The nature of traffic flow in a network depends upon application for which network is designed. Modeling traffic flow for a wireless sensor network (WSN) is as important as designing the MAC or routing protocol. In fact, estimating the flow of expected traffic is a prerequisite of designing a protocol. Efficient estimation of traffic flow helps in determining resource requirements. Overestimation and underestimation of the traffic pattern and flow can lead to wastage or exhaustion of resources. Medium access control (MAC) protocol is responsible for shared access of media among nodes. Poisson distribution is used for shaping of traffic for multilayer MAC (ML-MAC). This paper proposes another traffic profiles such as Pareto and generalized Pareto distribution (GPD) for ML-MAC which are more realistic than Poisson. Further, it simulates ML-MAC for the proposed traffic models. Finally, this paper compares the results in terms of energy consumption and average delay. The results are further justified and concluded by providing supporting applications.