Systems that Sustain Themselves: Energy Harvesting Sensor Nodes for Monitoring the Environment

Abstract

Battery-less sensor networks, that harvest energy from the ambient, have attracted much attention in the last few years due to the promise of low maintenance and untethered perpetual operation. However, the major challenge in such networks is that the availability of nodes in the network depends on the energy profile of their harvesting sources. This might affect network reliability. In this work, we study the suitability of Energy Harvesting Sensor (EHS) nodes, powered using light and vibrations, for a simple temperature monitoring application. We evaluate whether such an EHS node-based system can sustain itself and compare its performance with that of a traditional battery-based system. To economize on energy expenditure in the EHS system, we implement an Autoregressive (AR) model based adaptive sampling algorithm on the EHS nodes. After thorough experimental investigations, we conclude that the EHS node-based system fares quite well. Results show that adaptive sampling helps achieve energy savings of 62.12% and a 52.33% reduction in the amount of sampled data.