A distributed algorithm for approximate mobile sensor coverage

We consider the problem of covering a domain by mobile sensors and the design of an efficient schedule that reduces unnecessary sensor overlap and energy consumption. The problem is motivated by emerging participatory sensing applications as well as other coverage problems involving mobile nodes. The problem of minimizing the total energy consumption while maintaining the same level of coverage guarantee is NP-hard. We develop distributed algorithms achieving a constant approximation factor when sensors have unit disk sensing ranges, and a (1 + ε) approximation factor when the sensors also have constant bounded density. For all these algorithms the communication cost is asymptotically bounded by the cost of simply maintaining the direct neighborhood of each sensor. The constant approximation distributed algorithm can be generalized for the κ-coverage problem, when each point of interest has to be covered by at least κ sensors.