Skip to Main Content
Energy management is critical for emerging tiny wireless mobile devices such as sensors and PDAs, which are power limited by the capacity of their batteries. The energy cost of wireless transmission is high, being superlinear in the communication distance. Even an idle node that is turned on but otherwise not engaged in any activity may consume a significant amount of energy. Based on the observation that reduced communication distance between two communicating parties can significantly reduce the energy cost of the communication, we propose a motion prediction algorithm that exploits the movement history of a sender node to predict the likelihood that the node may move closer to the receiver by a given deadline. If the likelihood is high, then the transmission is postponed until the mobile node moves closer to the receiver at a future time. To reduce unproductive energy use by an idle node, we propose a sleep/wakeup algorithm that adoptively puts the node into sleep mode. However, while a mobile node is sleeping, it may miss critical opportunities for position update and motion prediction. The objective of this paper is then to realize an opportunistic scheduler that can balance the energy consumption by an idle system and the performance of motion prediction in saving energy for network transmission. The proposed system has been prototyped on an actual sensor network platform. Experimental results show that our opportunistic scheduler can achieve significant energy savings in terms of both network communication and keeping the system awake as necessary.