Distributed and adaptive power-aware scheduling of wireless sensor networks

In this paper, we propose a power-aware sleep scheduling scheme for a randomly deployed wireless sensor network comprising of sensors with varying footprints, akin to those of RF- or radar-based sensors. For this scheme, we first relate the performance of a sensor to its available power and then extend this relationship to the performance of the entire network. Then, to compensate for the effects of decrease in available power, we propose probabilistic scheduling controllers that maintain a desired probability of event detection. The key novelty in this work as compared to our previous work is that the proposed scheduling controller does not make any assumptions about the power decay model, which makes it suitable for practical implementation. Moreover, the proposed controller can also take care of increase in available power which is possible in practical systems with energy harvesting technologies. In addition to maintaining desired event detection probability, the proposed scheme is completely decentralized since it requires neither any global information nor any local communication for computing the control parameter at any time.