Using an energy aware adaptive scheduling algorithm to increase sensor network lifetime

• Main Author: Basford, Philip James
• Format: Thesis
• Language: English
• Published: University of Southampton 2015
• Subjects: glacier; Iceland
• Online Access: https://eprints.soton.ac.uk/375503/; https://eprints.soton.ac.uk/375503/1/__soton.ac.uk_ude_personalfiles_users_jo1d13_mydesktop_Basford.pdf

Sensor network research is an area which has experienced rapid growth in the last decade. One area in which it is proving particularly popular is that of environmental monitoring. Areas which have benefited from environmental monitoring include; volcanoes, crops, wildlife and, the test bed used for this thesis: glaciers. One of the main challenges faced by these networks is that of power management. This becomes even more important when energy harvesting techniques are used, as the availability of energy cannot be reliably predicted. In order to address this issue, an algorithm has been developed which allows a sensor node to adapt its schedule based on the available energy. This is achieved by using the average battery voltage to approximate energy reserves, then scaling the scheduled sensing tasks accordingly. This algorithm has been designed to work with differential GPS sensors which require multiple nodes to record in synchronisation. This means that a co-ordination system has been implemented to allow synchronisation between multiple systems with no direct communication methods. This thesis makes three main contributions to sensor network research: the development of a flexible platform for gateway nodes, the development and analysis of an energy aware adaptive scheduling algorithm, and algorithms for the use of alternate communication links to provide resilience in communications. Each of these contributions has been tested in Iceland as part of a real deployment to asses how they actually perform. During this deployment it has been possible to gather more data about the Skalafellsjokull glacier than has previously been achievable.