VulnerGAN: a backdoor attack through vulnerability amplification against machine learning-based network intrusion detection systems  

作　　者：Guangrui LIU Weizhe ZHANG Xinjie LI Kaisheng FAN Shui YU 

机构地区：[1]School of Cyberspace Science,Harbin Institute of Technology,Harbin 150001,China [2]Cyberspace Security Research Center,Peng Cheng Laboratory,Shenzhen 518055,China [3]School of Computer Science,University of Technology Sydney,Ultimo 2007,Australia

出　　处：《Science China(Information Sciences)》2022年第7期28-46,共19页中国科学（信息科学）（英文版）

基　　金：supported in part by National Key Research and Development Program of China(Grant No.2020YFB1406902);Key-Area Research and Development Program of Guangdong Province(Grant No.2020B0101360001);Shenzhen Science and Technology Research and Development Foundation(Grant No.JCYJ20190806143418198);National Natural Science Foundation of China(Grant No.61872110);Fundamental Research Funds for the Central Universities(Grant No.HIT.OCEF.2021007);Peng Cheng Laboratory Project(Grant No.PCL2021A02)。

摘　　要：Machine learning-based network intrusion detection systems(ML-NIDS) are extensively used for network security against unknown attacks. Existing intrusion detection systems can effectively defend traditional network attacks, however, they face AI based threats. The current known AI attacks cannot balance the escape rate and attack effectiveness. In addition, the time cost of existing AI attacks is very high. In this paper, we propose a backdoor attack called VulnerGAN, which features high concealment,high aggressiveness, and high timeliness. The backdoor can make the specific attack traffic bypass the detection of ML-NIDS without affecting the performance of ML-NIDS in identifying other attack traffic.VulnerGAN uses generative adversarial networks(GAN) to calculate poisoning and adversarial samples based on machine learning model vulnerabilities. It can make traditional network attack traffic escape black-box online ML-NIDS. At the same time, model extraction and fuzzing test are used to enhance the convergence of VulnerGAN. Compared with the state-of-the-art algorithms, the VulnerGAN backdoor attack increases33.28% in concealment, 18.48% in aggressiveness, and 46.32% in timeliness.

关 键 词：AI security adversarial sample data poisoning network intrusion detection generative adversarial network 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] TP393.08[自动化与计算机技术—控制科学与工程]