基于信息相关性的传感器网络数据聚合技术研究

项目名称： 基于信息相关性的传感器网络数据聚合技术研究

项目编号： No.61300239

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 自动化技术、计算机技术

项目作者： 郭剑

作者单位： 南京邮电大学

项目金额： 23万元

中文摘要： 无线传感器网络是以数据为中心的网络，任何应用都离不开数据的支撑。无线传感网目前存在着数据生产与数据消费不平衡的问题，对用户真正有用的信息被淹没在大量的感知数据之中。本课题拟针对此问题进行研究，以采集数据的信息相关性作为研究切入点，设计并实现符合传感网特征的数据聚合理论及方法。通过分析采集数据在时间上、空间上，多维数据之间以及多视角图像、视频之间的相关性，使用相关性分析、预测算法、地学统计、空间数据挖掘、关联分析、多视角分布式编解码等多种技术手段，探寻冗余数据的消除方法，设计能量高效的聚合方案，提升采集数据的信息质量。在此基础之上，结合节点调度、分布式存储等技术手段，从数据聚合视角下对拓扑优化控制与数据存储查询展开研究。以上述研究为支撑，设计并实现一个面向水环境监测的无线传感器网络数据聚合原型系统。

中文关键词： 无线传感器网络；数据聚合；信息相关性；网络优化；

英文摘要： Wireless sensor networks are data-centric networks, and no application is continuable without data. However, there is a big problem of imbalance between data generation and data consumption in wireless sensor networks, which makes useful information drowned in the numerous sensory data. Aiming at this problem, this project tries to take the information correlation as the starting point, and to design and realize a series of theories and methods of data aggregation for wireless sensor networks. By analyzing the correlation of perceptual data on time and space, the correlation of multi-dimensional data, and the correlation of multi-view image and video, our purpose is to explore the methods of eliminating redundant data with different technology such as interdependency analysis, prediction algorithm, geostatistics, spatial data mining, correlation analysis, multi-view distributed encoding. Based on this, we will design energy-efficient aggregating schemes to improve the information quality of sensory data. And from the view of data aggregation, we will have research on topology optimizing and data storage and query through node scheduling and distributed storage technique. At last, based on the above researches, we will design and realize a data aggregation prototype system of wireless sensor networks for water en

英文关键词： wireless sensor networks；data aggregation；information correlation；network optimization；