Conserving Energy in Dense Sensor Networks via Distributed Scheduling (Extended Abstract)

In wireless sensor networks, sensors are often batterypowered and cannot be recharged. Even the simple act of sensing its environment can consume significant amounts of these devices’ energy. To reduce the likelihood that an insufficient number of sensors are deployed within a region, sensors will often be deployed at a greater density than what is expected to be necessary. The sensing devices can utilize this increased density to conserve energy by time-sharing the sensing load needed to cover their respective sensing region. Ideally, to maximize sensor lifetimes, sensors that cover a shared region can coordinate their responsibility for covering that region with a schedule that has all but one of the sensors turned off at any given time within that region. What complicates this scheduling, however, is the fact that neighboring regions are covered by non-identical but overlapping sets of sensor devices. In this paper, we perform an initial investigation into the problem of scheduling the on periods of sensor nodes in densely covered regions. Our goal is to reduce energy consumption per sensor while maintaining sufficient coverage of all regions that the sensors were deployed to cover. Few works have considered the problem of scheduling among multiple devices in dense regions. [4] and [3] are perhaps most closely related to our work here, both of which deal with energy-coverage problem in sensor networks. They both use probablistic probing schemes to determine when a sensor can safely go to sleep without losing coverage. Our work differs in that sensor coverage times are pre-determined, obviating the need for sensors to participate in a polling process that uses valuable energy to trasmit and receive polling messages. We introduce our problem formulation in next section, describe our scheduling algorithm and prove some of its basic desirable properties in Section 3, and conclude in Section 4.