Adaptive sleep discipline for energy conservation and robustness in dense sensor networks

This paper presents an adaptive approach for conserving energy in high-density sensor networks. The proposed method allows sensor nodes to sleep while ensuring that application performance constraints are met. Unlike deterministic algorithms that assume static connectivity, the approach uses a randomized algorithm to provide robustness to the variations in network connectivity. These variations are due to fading channels, depletion or addition of nodes, node mobility, and the sleeping of nodes. The algorithm developed in the paper is extremely lightweight and does not require nodes to keep any state information about their individual neighbors. Based on local observations, each node independently decides when to sleep and wakeup. The algorithm also ensures an evenly distributed workload among nodes, and achieves energy savings proportional to the density of nodes.