无线传感网络中的自适应入侵检测算法  

Adaptive intrusion detection in wireless sensor networks

作　　者：苏明[1] SU Ming(Beijing Open University,Beijing 100018,China)

机构地区：[1]北京开放大学,北京100081

出　　处：《导航定位学报》2020年第4期106-110,共5页Journal of Navigation and Positioning

基　　金：北京市教委-自然科学基金重点项目(KZ201951160050)。

摘　　要：针对无线传感网络的入侵检测算法因大多基于数据挖掘算法,所建立的监测系统易遭受未知攻击的问题,提出1种自适应的入侵检测算法:跟踪每个子系统的接收操作特征(ROC),再依据ROC行为特征,自适应地调整转发至每个子系统的融合数据的比例,达到在数据融合阶段运行入侵检测系统,实现有效监测的目的。仿真结果表明,基于该算法的监测系统具有99%的准确率。Aiming at the problem that it is liable to unknown attack for the monitoring and detection system established by the intrusion detection algorithm of wireless sensor networks mostly based on data-mining method,the paper proposed an adaptive intrusion detection algorithm:the receiver operating characteristics(ROC)of each subsystem were tracked,and based on the improvement/degradation of the ROC behavior,the proportion of aggregated data of each subsystem forwarded was adaptivley adjusted,which leads to the efficient monitoring and detection by implementing the intrusion detection in the data fusion phase.Simulational result showed that the sytem with the proposed algorithm could provide up to an accuracy rate of 99%.

关 键 词：无线传感网络 入侵检测 机器学习 簇 接收操作特征 

分 类 号：TPT393[自动化与计算机技术]