Distributed Multi-Pipeline Scheduling with Low Collisions in Low-Duty-Cycle Sensor Networks with Unreliable Communication Links
Chan Jiang, Junbin Liang and Taoshen Li
Abstract: Low-duty-cycle sensor networks (LDC-SNs) are a new type of wireless sensor networks (WSNs). Similar with traditional WSNs, LDC-SNs are composed of large amounts resource-limited nodes, and they are deployed in harsh environment to perform long-time surveillance tasks. However, they have a new feature that their nodes would sleep for most of their time in each working period to save energy. The sleeping will cause serious latency in data collection. In order to reduce the latency, existing works adopt multi-pipeline scheduling to adjust the nodes’ wakeup time, i.e., the nodes in multiple data transmission paths are waked up in turn so that they can relay the data quickly. However, these works seldom consider transmission collisions and energy balance, so they have heavy data retransmission burden and the network lifetime is short. To solve the problem, we proposed a distributed and energy-balanced multi-pipeline scheduling algorithm with low collisions. The algorithm considers factors such as communication quality, energy levels of the nodes and load balance of communications, in the process of multi-pipeline scheduling. Theoretical analyses and simulations show that the algorithm can not only shorten the data collection latency, but also balance energy consumption of nodes to extend the network lifetime.
Keywords: Low-duty-cycle sensor networks; distributed algorithm; multi-pipeline scheduling; data collection