北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2012, Vol. 35 ›› Issue (1): 41-45.doi: 10.13190/jbupt.201201.41.xionga

• 论文 • 上一篇    下一篇

节点信誉相关的无线传感器网络故障检测

熊翱,赵晓东,高志鹏,黄日茂,郭全
  

  1. 北京邮电大学 网络与交换技术国家重点实验室, 北京 100876
  • 收稿日期:2011-10-18 修回日期:2011-10-24 出版日期:2012-02-28 发布日期:2012-01-05
  • 通讯作者: 赵晓东 E-mail:xiyangxiaodong@126.com
  • 作者简介:熊翱(1974-),男,副教授,硕士生导师,E-mail:xiongao@bupt.edu.cn
  • 基金资助:

    国家自然科学基金项目(60821001,60802035,60973108)

Wireless Sensor Networks Fault Detection via Node Credit Value

XIONG Ao,ZHAO Xiao-dong,GAO Zhi-peng,HUANG Ri-mao,GUO Quan
  

  1. State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications, Beijing 100876, China
  • Received:2011-10-18 Revised:2011-10-24 Online:2012-02-28 Published:2012-01-05
  • Contact: Xiao-Dong ZHAO E-mail:xiyangxiaodong@126.com

摘要:

提出一种基于节点信誉的无线传感器网络故障检测算法.算法为每个传感器节点建立了基于Dirichlet分布的信誉函数,整个算法由检测模块、信誉更新模块和故障判定模块3部分组成.检测模块计算节点当前时刻的健康等级;信誉更新模块负责维护、更新节点的信誉值;故障判定模块则对是否发生节点故障和敏感事件进行判定.仿真结果表明,该算法具有较高的故障检测精度,在传感器网络故障节点达到25%的情况下,仍能检测出9483%的故障节点,而误警率则可以保持在5%以下.

关键词: 无线传感器网络, 故障检测, 信誉, Dirichlet分布

Abstract:

A fault detection algorithm based on Dirichlet reputation for wireless sensor networks was proposed. The algorithm consists of three modules: measurement, reputation updating and fault judgment. The healthy level of sensor node in current time was evaluated in the measurement module. The work of maintaining and updating the reputation of the sensor node were performing by the reputation updating module. Fault judgment module determines whether a node is faulty or not, and whether the sensitive event happens or not. Simulation results show that the proposed algorithm has a high fault detection rate. Even 25 percent of nodes are faulty, 94.83 percent of fault nodes can be detected out and the false alarm rate is below 5%.

Key words: wireless sensor network, fault detection, reputation, Dirichlet distribution

中图分类号: