Network Connectivity under Probabilistic Communication Models in Sensor Networks

Several previous works have experimentally shown that communication ranges of sensors are not regular disks. Rather, they follow probabilistic models. Yet, many current connectivity maintenance protocols assume the disk communication model for convenience and ease of analysis. In addition, current protocols do not provide any assessment of the quality of communication between nodes. In this paper, we take a first step in designing connectivity maintenance protocols for more realistic communication models. We propose a distributed connectivity maintenance protocol that explicitly accounts for the probabilistic nature of communication links and achieves a given target communication quality between nodes. Our protocol is simple to implement, and we demonstrate its robustness against random node failures, inaccuracy of node locations, and imperfect time synchronization of node clocks using extensive simulations. We compare our protocol against others in the literature and show that it activates fewer number of nodes, consumes much less energy, and significantly prolongs the network lifetime. We extend our connectivity protocol to provide probabilistic coverage as well, where the sensing ranges of sensors follow probabilistic models.