Sleep scheduling of wireless sensor networks using hard-core point processes

Abstract

In this paper, we examine a randomly deployed sensor network and propose a probabilistic sleep-scheduling scheme to conserve energy while maintaining a desired level of coverage by using coordination among sensors. To introduce coordination among neighboring sensors, we use a hard-core point process from stochastic geometry, which is an inhibition process and does not allow more than one sensor to be on in a given circular region. As compared to existing schemes, the hard-core point process allows coordination between sensors with little communication overhead. We analyze the proposed scheme and derive an expression for the event detection probability. We also perform Monte Carlo simulations and compare the event detection probability to a random switching scheme for a Poisson process of same intensity in order to confirm that the proposed scheme improves the detection probability.