Wireless sensor networks can be invaluable in a diverse range of applications spanning both military and civilian sectors. A distributed, ad-hoc wireless sensor network consists of many hundred sensor nodes scattered throughout an area of interest. The sensor nodes are designed specially for distribution in adverse and inaccessible areas without fixed infrastructure. Each node contains both communication and processing elements and is designed to monitor some aspect of its environment. Wireless sensor nodes are traditionally battery driven and therefore operate on an extremely frugal energy budget. To avoid the need for periodic battery replacement, there is a desire to increase the lifetime of the sensor nodes through intelligent power generation and management. The aim of this project is to design and build an energy-aware wireless sensor network using both hardware and embedded software techniques.

Many distributed sensor applications sample their environment then forward the data to a central location on a periodic basis. In general, the energy consumption of a node is dominated by the communication subsystem, for example simply turning on a commercially available RF transceiver is high. However, designing circuits and systems that consume less power is technically challenging. Thus, innovative solutions in transceiver and protocol design are required to achieve energy-efficient transmissions.

Starting with a basic sensor module built around a 433MHz RF transceiver and a low-cost 8-bit microcontroller, the group will design and build a solar power supply unit that will provide the basis of an energy efficient power management and networking scheme. Energy savings will be realized by implementing media-access control and routing protocols that carefully manage their use of communications. Each sensor node will remain in a sleep state until awoken by either an internal alarm, external event or a request for data from another node. A duty-cycle approach to communications will be taken, thus requiring that each node exchange sleep-and-listen schedules and maintain synchronous clocks to a certain degree of accuracy. By providing a means to measure the battery status of each node, a simple energy-aware routing technique will also be investigated.

Kelvin Nicholson, Associate Professor Graeme Pendock, Clare Kitching