Time-Sensitive Utility-Based Single-Copy Routing in Low-Duty-Cycle Wireless Sensor Networks
Time-Sensitive Utility-Based Single-Copy Routing in Low-Duty-Cycle Wireless Sensor Networks
复制标题
DOI：
10.1109/tpds.2014.2321136
复制
发表时间：
2015-05
影响因子：
5.3
通讯作者：
Mingjun Xiao;Jie Wu;Liusheng Huang
中科院分区：
文献类型：
--
作者：
Mingjun Xiao;Jie Wu;Liusheng Huang
基于效用的路由是一种基于特殊复合效用度量的路由方案。现有的基于效用的路由算法还没有考虑投递延迟，不能很好地适应低占空比的无线传感器网络。在本文中，我们提出了一种时间敏感的实用模型。成功的端到端消息传递将获得积极的好处，随着传递延迟的增加而线性减少；否则，失败的传递将获得零好处。无论消息传递成功还是失败，效用都是收益减去总传输成本。这种实用新型类似于现实世界中的邮政服务。在这种新的实用模型下，我们分别针对非重传设置和允许重传设置设计了两种时间敏感的基于效用的最优路由算法。在我们的设计中，我们推导了一个迭代公式来计算每次消息传递的期望效用，并提出了一种二进制搜索方法来确定最优的重传次数。结果表明，这两种算法都能达到每次消息传递的最优期望效用，这是效益、可靠性、延迟和成本等相关因素之间的最优平衡。仿真结果也证明了所提算法的显著性能。
Utility-based routing is a routing scheme based on a special composite utility metric. The existing utility-based routing algorithms have not yet considered the delivery delay, so that they cannot work well in low-duty-cycle wireless sensor networks (WSNs). In this paper, we present a time-sensitive utility model. A successful end-to-end message delivery will obtain a positive benefit, which linearly decreases along with an increasing delivery delay; otherwise, a failed delivery will receive zero benefit. The utility is the benefit minus the total transmission costs, no matter if the message delivery succeeds or fails. Such a utility model is analogous to the postal service in the real world. Under this novel utility model, we design two optimal time-sensitive utility-based routing algorithms for the non-retransmission setting and the retransmission-allowed setting, respectively. In our designs, we derive an iterative formula to compute the expected utility of each message delivery, and we present a binary search method to determine the optimal retransmission times. As a result, the two algorithms can achieve the optimal expected utility for each message delivery, which is the optimal balance among the concerned factors, including benefit, reliability, delay, and cost. The simulation results also prove the significant performances of our proposed algorithms.