Energy-Efficient Stochastic Target Coverage in Sensor Surveillance Systems

In many sensor surveillance systems such as environment and habitat monitoring, sensors are deployed in an area to monitor several important locations where targets appear. As the sensor nodes are often powered by non-rechargeable batteries, it is crucial to extend the lifetime of sensor surveillance systems. How to cover multiple random targets and extend the system lifetime at the same time is a challenge. Previous studies show that data fusion improve coverage by exploiting the collaboration among sensors. However, they assume that targets deterministically appear at some specific locations and do not consider the randomness of the appearance of the targets. Considering both the high cost of the sensor surveillance system and various reliability requirements in practice, we explore the tradeoff between the system lifetime and the system reliability in this paper. We formulate the stochastic target coverage problem as probabilistic set covering. After showing its NP-Completeness, we design a PSC-based heuristic algorithm in the data fusion model. Extensive simulations show that the system lifetime can be extended dramatically with a small loss to the system reliability.