Adversarial Deep Ensemble: Evasion Attacks and Defenses for Malware Detection

被引:75
|
作者
Li, Deqiang [1 ]
Li, Qianmu [1 ]
机构
[1] Nanjing Univ Sci & Technol, Sch Comp Sci & Engn, Nanjing 210094, Peoples R China
关键词
Adversarial Machine Learning; Deep Neural Networks; Ensemble; Adversarial Malware Detection;
D O I
10.1109/TIFS.2020.3003571
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
Malware remains a big threat to cyber security, calling for machine learning based malware detection. While promising, such detectors are known to be vulnerable to evasion attacks. Ensemble learning typically facilitates countermeasures, while attackers can leverage this technique to improve attack effectiveness as well. This motivates us to investigate which kind of robustness the ensemble defense or effectiveness the ensemble attack can achieve, particularly when they combat with each other. We thus propose a new attack approach, named mixture of attacks, by rendering attackers capable of multiple generative methods and multiple manipulation sets, to perturb a malware example without ruining its malicious functionality. This naturally leads to a new instantiation of adversarial training, which is further geared to enhancing the ensemble of deep neural networks. We evaluate defenses using Android malware detectors against 26 different attacks upon two practical datasets. Experimental results show that the new adversarial training significantly enhances the robustness of deep neural networks against a wide range of attacks, ensemble methods promote the robustness when base classifiers are robust enough, and yet ensemble attacks can evade the enhanced malware detectors effectively, even notably downgrading the VirusTotal service.
引用
收藏
页码:3886 / 3900
页数:15
相关论文
共 50 条
  • [1] Adversarial superiority in android malware detection: Lessons from reinforcement learning based evasion attacks and defenses
    Rathore, Hemant
    Nandanwar, Adarsh
    Sahay, Sanjay K.
    Sewak, Mohit
    [J]. FORENSIC SCIENCE INTERNATIONAL-DIGITAL INVESTIGATION, 2023, 44
  • [2] Adversarial superiority in android malware detection: Lessons from reinforcement learning based evasion attacks and defenses
    Rathore, Hemant
    Nandanwar, Adarsh
    Sahay, Sanjay K.
    Sewak, Mohit
    [J]. FORENSIC SCIENCE INTERNATIONAL-DIGITAL INVESTIGATION, 2023, 44
  • [3] Robust Malware Detection Models: Learning from Adversarial Attacks and Defenses
    Rathore, Hemant
    Samavedhi, Adithya
    Sahay, Sanjay K.
    Sewak, Mohit
    [J]. FORENSIC SCIENCE INTERNATIONAL-DIGITAL INVESTIGATION, 2021, 37
  • [4] Defending malware detection models against evasion based adversarial attacks
    Rathore, Hemant
    Sasan, Animesh
    Sahay, Sanjay K.
    Sewak, Mohit
    [J]. PATTERN RECOGNITION LETTERS, 2022, 164 : 119 - 125
  • [5] PAD: Towards Principled Adversarial Malware Detection Against Evasion Attacks
    Li, Deqiang
    Cui, Shicheng
    Li, Yun
    Xu, Jia
    Xiao, Fu
    Xu, Shouhuai
    [J]. IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING, 2024, 21 (02) : 920 - 936
  • [6] Adversarial Human Context Recognition: Evasion Attacks and Defenses
    Alajaji, Abdulaziz
    Gerych, Walter
    Chandrasekaran, Kavin
    Buquicchio, Luke
    Agu, Emmanuel
    Rundensteiner, Elke
    [J]. 2023 IEEE 47TH ANNUAL COMPUTERS, SOFTWARE, AND APPLICATIONS CONFERENCE, COMPSAC, 2023, : 223 - 232
  • [7] Adversarial Attacks and Defenses in Deep Learning
    Ren, Kui
    Zheng, Tianhang
    Qin, Zhan
    Liu, Xue
    [J]. ENGINEERING, 2020, 6 (03) : 346 - 360
  • [8] Automated poisoning attacks and defenses in malware detection systems: An adversarial machine learning approach
    Chen, Sen
    Xue, Minhui
    Fan, Lingling
    Hao, Shuang
    Xu, Lihua
    Zhu, Haojin
    Li, Bo
    [J]. COMPUTERS & SECURITY, 2018, 73 : 326 - 344
  • [9] Adversarial Attacks on Mobile Malware Detection
    Shahpasand, Maryam
    Hamey, Len
    Vatsalan, Dinusha
    Xue, Minhui
    [J]. 2019 IEEE 1ST INTERNATIONAL WORKSHOP ON ARTIFICIAL INTELLIGENCE FOR MOBILE (AI4MOBILE '19), 2019, : 17 - 20
  • [10] Adversarial Attacks and Defenses for Deep Learning Models
    Li, Minghui
    Jiang, Peipei
    Wang, Qian
    Shen, Chao
    Li, Qi
    [J]. Jisuanji Yanjiu yu Fazhan/Computer Research and Development, 2021, 58 (05): : 909 - 926