A monitoring lifetime metric for sensor networks

Sensor networks are intended to provide continuous coverage to monitored areas and to communicate their sensed information back to designated gateway nodes. Since sensors are battery-powered and the lifetime of each node is finite, the monitoring and communication functions compete for the same energy resources. It has been established that for sensor networks the goal of coverage-aware (as distinct from merely energy-aware) communications protocols should be to maximize the monitoring lifetime while ensuring delivery of information. This requires that information be routed in ways that uniformly deplete the monitoring lifetime across the area of interest. A new metric for monitoring lifetime has been developed using a Markov random field model. The metric is based on a combination of remaining battery life and the spatial distribution of nodes to account for overlapping coverage. The result is a true coverage availability measure. Calculation of the metric uses a constraint-based clustering technique which is distributed and efficient, and therefore may be implemented by each node using information reported only from its neighbors. Simulation results show that this method provides a reliable indicator of actual waning monitoring lifetime. This suggests that the method may be used in conjunction with control techniques for improved coverage balancing, and indeed coverage awareness.