高效容错的无线传感网事件及其边界检测算法

发布时间：2019-02-12 16:56

【摘要】：事件检测与事件边界检测是无线传感器网络的重要应用之一,节点故障的准确检测是提高事件与事件边界检测效率的前提.然而,目前的故障检测机制对节点故障类型分析不够明晰,导致系统可能将事件边界节点误判为故障节点,且常需要传感器节点间进行频繁通信,导致网络系统容错性能和节点利用率低下,并带来额外的能耗开销.为了达到较高的检测精度与能源利用率,提出了一种新的高效容错的无线传感网事件及其边界检测算法:利用时间相关性实现无线传感器网络事件检测,利用空间相关性实现故障检测与事件边界检测;提出了节点的信息可靠度恢复机制,使得节点能够根据网络环境的变化,自动调整节点的信息可靠度.实验结果表明,即使在故障概率较高的情况下,该策略仍然具有良好的性能表现.
[Abstract]:Event detection and event boundary detection is one of the important applications in wireless sensor networks. The accurate detection of node faults is the prerequisite to improve the efficiency of event and event boundary detection. However, the current fault detection mechanism is not clear enough to analyze the node fault type, which leads to the system may misjudge the event boundary node as the fault node, and often requires frequent communication between sensor nodes. It leads to low fault-tolerant performance and node utilization in network systems, and brings additional energy consumption overhead. In order to achieve higher detection accuracy and energy efficiency, a novel and efficient fault tolerant event detection algorithm for wireless sensor networks and its boundary detection is proposed. Fault detection and event boundary detection are realized by spatial correlation. The information reliability recovery mechanism of nodes is proposed, which enables nodes to automatically adjust the information reliability of nodes according to the changes of network environment. The experimental results show that the strategy has good performance even when the fault probability is high.
【作者单位】：南京邮电大学计算机学院;计算机软件新技术国家重点实验室(南京大学);江苏省无线传感网高技术研究重点实验室(南京邮电大学);宽带无线通信与传感网技术教育部重点实验室(南京邮电大学);
【基金】：国家自然科学基金项目(61202004) 中国博士后科学基金项目(2011M500095,2012T50514) 江苏省博士后科研资助计划项目(1102103C) 江苏省自然科学基金项目(BK2011754,BK2009426) 江苏省高校自然科学研究计划项目(12KJB520007) 南京大学计算机软件新技术国家重点实验室基金项目(KFKT2012B21)
【分类号】：TP212.9;TN929.5

【参考文献】