WSNs中带有置信度和误差限的序列匹配算法  

Time Series Marching Algorithm with Confidence and Error Bounds in WSNs

作　　者：王甜[1] 董渭清[1] 王鹏[1] 

机构地区：[1]西安交通大学电子与信息工程学院,陕西西安710049

出　　处：《小型微型计算机系统》2008年第6期1027-1030,共4页Journal of Chinese Computer Systems

基　　金："九八五行动计划"基金项目资助

摘　　要：在分析WSNs数据特点的基础上,提出WSNs中基于置信度和误差限的序列匹配算法.定义数据序列的距离和基于置信度和误差限的匹配,并利用距离计算不同匹配度的节点.采用自底向上的线性拟合对数据进行处理解决数据的不对齐性,丢失性等对序列匹配带来的问题.使用真实传感器网络数据对算法进行测试,分析试验结果证实采用基于置信度和误差限的序列匹配算法的WSNs比普通WSNs更加节约能耗.This paper developed a time series marching algorithm based on confidence and error bounds in WSNs. The algorithm defines distance between time series and the matching degree based on confidence and error bounds, computes how time series match according to the distance. Bottom-up linear approximation is used to resolve the problem of data losing and drifting. Experiment result shows that WSNs with time series marching algorithm can save more energy than usual WSNs.

关 键 词：传感器网络 序列匹配 置信度 误差限 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]