A Distributed Policy Scheduling for Wireless Sensor Networks

We aim to save energy by scheduling periodic sensors\´ duty cycles. Different from existing works, we integrate two important tasks, collision avoidance and duty cycling, into one scheduling of sensors\´ activities. Our scheme consists of two parts: a coloring scheme that assigns a "color" to each sensor and a color scheduling scheme that associates each color with a schedule. Each sensor switches between sleep and active modes according to the schedule associated with its color. We propose a coloring definition. Theoretical analyses and simulation results show that, if this coloring is used by collision avoidance schemes such as TDMA or CDMA, connectivity can be guaranteed by using much less colors than traditional colorings. Theoretical analyses and simulation evaluations are presented for our scheduling. As compared to handling collision avoidance and duty cycling separately, performances can be improved by integrating these two tasks into one carefully designed schedule of sensors\´ activities. In particular, in duty cycled dense networks, in order to reduce packet latency and save energy, it is desirable to guarantee the communication connectivity of links only in a sparse connected subgraph.