高效物联网虚假数据注入攻击智能防御仿真  

Intelligent Defense Simulation of High Efficient Internet of Things False Data Injection Attack

作　　者：王艳丽[1] 吕海翠[1] 宋佳[1] WANG Yan-li;LV Hai-cui;SONG Jia(Haojing College of Shaanxi University of Scienceand&Technology,Xi'an Shanxi 712046,China)

机构地区：[1]陕西科技大学镐京学院,陕西西安712046

出　　处：《计算机仿真》2020年第9期258-261,337,共5页Computer Simulation

基　　金：陕西省教育厅项目(16JK2048)。

摘　　要：为了使用户能够更为安全地操作物联网系统,提出了一种新的高效物联网虚假数据注入攻击智能防御方法。根据虚假数据注入攻击原理,获取其运行方式与攻击手段,利用监督保局映射方法选择注入攻击特征,融合支持向量机理论对物联网数据的特征信息量进行真伪信息筛选与聚类,进而完成对虚假数据的特征检测,在此基础上,构建三层动态攻防博弈模型,模拟虚假数据注入攻击与对应传统防御方法的所有运作流程,并对模型进行求解,从而得到最优的防御方案。仿真结果证明,所提方法能够更为有效的对虚假数据注入攻击进行防御,并且防御时间较短。In order to enable users to operate the Internet of things system more safely,an efficient intelligent de⁃fense method of false data injection attack in the Internet of things was put forward.According to the principle of false data injection attack,the operation mode and attack means were obtained.Moreover,the injection attack features were selected by the supervised locality preserving projection.Combined with the support vector machine theory,the feature information of IoT data was screened and clustered,and then the feature detection of false data was completed.On this basis,a three-layer dynamic attack and defense game model was constructed to simulate all the operation processes of false data injection attacks and traditional defense methods.Meanwhile,the model was solved to get the optimal defense scheme.Simulation results show that the proposed method can effectively protect against false data in⁃jection attacks,and the defense time is shorter.

关 键 词：虚假数据 智能防御 物联网 特征提取 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]