Light-Weight Energy Consumption Model and Evaluation for WirelessSensor Networks

Abstract

Wireless sensor networks are comprised of low power devices with fixed energy stores. They often require long term operation for successful deployment so it is important to efficiently manage and track their energy usage. To effectively accomplish this across distributed networks requires methods which have low energy cost with minimal error. In this paper we present a straightforward model for energy consumption in wireless sensor networks which is light-weight and accurate. The model has been applied to a wireless sensor network developed by the Queen's University MEMs lab and is evaluated with a custom testbed. Through testing, the model is exposed to realistic disturbances of communication loss, battery effects and variable voltage supplies. It was shown that with 99% packet reception rates in the network, the model accurately estimates end node energy consumption with less than 5% error. These results were demonstrated across varying data rates, battery supply capacities, and runtimes up to full network lifetime.