WSN中事件驱动与信任分配加权的层次数据融合  

Hierarchical Data Fusion Using Event-Driven and Belief Assignment in WSN

作　　者：余修武[1,2,3] 张枫 周利兴[1,2] 刘永[1,2,3] 刘琴 雷林[1] 汪弘 YU Xiuwu1＇2＇3 , ZHANG Feng1,2 ZHOU Lixing1,2 LIU Yong1,2,3, LIU Qin1＇2 ,LEI Lin1, WANG Hong1(1. University of South China ,Environmental protection and safety engineering institute ,Hengyang Hu＇naa 421001, China ; 2.State Key Laboratory of Safety and Health for Metal Mines,Maanshan Anhui 243000, China; 3.Hunan Engineering Research Center for Uranium Tailings Decommission and Treatment,Hengyang Hu＇nan 421001 ,Chin)

机构地区：[1]南华大学环境保护与安全工程学院,湖南衡阳421001 [2]金属矿山安全与健康国家重点实验室,安徽马鞍山243000 [3]湖南省铀尾矿库退役治理技术工程技术研究中心,湖南衡阳421001

出　　处：《传感技术学报》2017年第12期1948-1953,共6页Chinese Journal of Sensors and Actuators

基　　金：金属矿山安全与健康国家重点实验室开放基金项目(2016-JSKSSYS-04);湖南省教育厅科研重点项目(15A161);江西省自然科学基金项目(20122BAB201050)

摘　　要：针对WSN监测突发事件及节点能量受限问题,提出了一种基于事件驱动与信任度分配加权的层次数据融合算法(EDBA)。通过设置监测阈值,仅在事件发生时,相关部分节点才进入高频次数据采集和传输的兴奋状态,其他情况节点处于低频次采集(或传输)的抑制(或活动)状态,采用证据理论及信任分配函数对网络监测数据进行多层次融合,以减少监测数据传输量。仿真表明,在通常情况下,EDBA算法能耗分别是EBPDF、LEACH的50%和21%,有效地降低了网络能耗。In view of unexpected events and node energy limitation in WSN,a data fusion algorithm using eventdriven and belief assignment（ EDBA） is proposed.By setting monitoring threshold,only when the unexpected events occurs,the relevant nodes switching to excited state which can collect and transfer data in high frequency,in other cases,nodes keeping suppressed or active state to collect and transfer data.Monitoring data is fused through evidence theory and belief assignment function to reduce the amount of transmitted data.Simulation and analysis show that under normal circumstances,energy consumption of EDBA algorithm are respectively 50% of EBPDF and 21% of LEACH.EDBA can effectively save network energy.

关 键 词：无线传感器网络 层次融合 事件驱动 信任分配 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]