Modeling and Performance Evaluation of Energy Consumption in S-MAC Protocol Using Generalized Stochastic Petri Nets

Abstract

One of the specifications of wireless sensor networks is the limited power sources for their nodes. Therefore, assessment of energy consumption in these networks is very important, andusing traditional simulators to evaluate the energy consumption of network nodes has become common practice. Simulators often have problems such as fluctuating output values in different implementations of the same scenario, lack of a visual structure to understand the system, and lack of theoretical background for the gained results. In contrast, analytical modeling methods such as Petri nets do not have the aforementioned problems. Also, they have the necessary tools to model and evaluate the performance of investigated system.The Sensor-MAC (S-MAC) protocol in these networks, which operates in MAC layer, is a competition-based protocol designed to reduce power consumption. This paper presents an analytical model based on a Generalized Stochastic Petri Net (GSPN) to evaluate the power consumption of nodes in an S-MAC-based wireless sensor network. The presented Petri model was implemented in PIPE software, and by applying mathematical calculations from the model, we were able to derive the equations for calculating its energy consumption. The validity of the proposed model is measured by implementing similar scenarios in the Castalia simulator. The experiments were conducted in terms of the number of nodes, duty cycle rate, upper layer data flow, and packet size. The results of the presented model are extracted in a much shorter time than the simulator with the same gained values.