Cool: On coverage with solar-powered sensors

In this paper, we study the dynamic node activation schedule for the utility based coverage problem in solar-powered wireless sensor networks. We assume that the utility achieved by a WSN for coverage service is a submodular function over the set of sensors that will provide the service. We first present an integer programming formulation with submodular objective functions.We then present an efficient simple greedy hill-climbing algorithm such that the achieved average utility of the computed schedule is at least 1/2 times that achieved by the optimal schedule. To the best of our knowledge, this is the first polynomial time algorithm that can ensure a good constant approximation of the achieved utility for multi-target coverage problem. We conduct extensive evaluations to study the performances of our proposed aggregation scheduling algorithm on real testbed. Our evaluation results corroborate our theoretical analysis.