无线传感网中事件监测的混沌表示与异常识别

项目名称： 无线传感网中事件监测的混沌表示与异常识别

项目编号： No.61271274

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 无线电电子学、电信技术

项目作者： 王典洪

作者单位： 中国地质大学（武汉）

项目金额： 73万元

中文摘要： 监控环境或对象中是否有异常事件发生是无线传感网（WSN）的核心应用目标之一。由于事件的多样性、复杂性和传感器节点监测数据的不确定性，使得现有的WSN事件监测系统还停留在短期的小规模应用水平上。本项目将以综合优化WSN事件监测的节能性、实时性和可靠性为目标，为大规模长期部署的WSN事件监测提供一系列新方法。主要包括：（1） 采用改进的符号聚合近似方法，构建高效节能的空时数据序列压缩与融合机制；（2）提出一种WSN事件监测的混沌表示模型，形式化定义和准确刻画复杂事件的空时变化信息，利用混沌表示模型的统计信息和分形特征，sink网关构建监测区域的全局事件视图，传感器节点分布式地构建监测区域的局部事件视图；(3)在全局和局部事件视图上，寻求轻量级、在线的异常识别算法，以期能高效准确地检出异常事件，从而满足实际的预警应用需求。

中文关键词： 无线传感器网络；数据压缩；数据融合；异常事件检测；轻量级算法

英文摘要： The core application goal of the wireless sensor networks (WSN) is to detect whether an abnormal event happens in the monitoring environment or object. Due to the event diversity and complexity and the uncertainty of the sensor nodes monitoring data, the existing WSN event monitoring system is still stuck in the level of short-term and small-scale applications. The project, which regards optimizing energy efficiency, real-time and reliability of WSN event monitoring as a target, will provide a range of new methods for large-scale and long-term deployment of WSN event monitoring. The methods include: (1)Using an improved symbolic aggregate approximation method to build the energy-efficient space-time data compression and fusion mechanism;(2)Proposing a chaos representation model of WSN event monitoring to define and depict the space-time change of the complex event information, and then the sink gateway can build the global event view of monitoring region while sensor nodes can distributed build the local event view of monitoring region by using the statistics and fractal characteristics of the chaos representation mode;(3)In view of global and local events, seeking the lightweight and online anomaly identification algorithm which can detect abnormal events efficiently and accurately to meet the actual demand of

英文关键词： wireless sensor networks；data compression；data fusion；abnormal event detection；lightweight methods