Performance Analysis of a Dynamically Power Managed and Event-Based Wireless Sensor Node Enabled by Queue Discipline

In recent years, Wireless Sensor Networks (WSNs) have attracted the attention of researchers in view of providing a system with high performance and low power consumption.The consistent challenge is to address the trade-off between performance and power consumption.Hence, many key performance metrics need to be analysed for the design of an efficient Wireless Sensor Network.Existing power management techniques, when surveyed, have addressed the issue of power and performance of the node but with the limitation on the selection of queue discipline; this motivates our study to analyse the importance on the selection of queue discipline as it majorly plays a vital role in power and performance management.Thus, we developed a Dynamically Power Managed WSN node in MATLAB Simulink depicting the stochastic behaviour of event arrival and performed the analysis on a single server.This article focuses on the study of queue discipline based on M/M/1 queuing theory with a detailed analysis of First In First Out Queue on the performance of an individual WSN node.The innovation of our work is in the detailed analysis of the behaviour of events in the queue.The parameters analysed are waiting time, average length of the event in the queue and the number of events missed the service by the server.The simulation results make it evident that the events when served based on the First Come First Serve basis performs 10% better in terms of missing events in the queue as compared to Last in First Out Discipline.It is also observed that the arrival rate of the events has an impact on the number of missed events and the utilization of the processor; hence we analysed the need to reduce the number of missed events, especially when the arrival of events is fast.