Improve the robustness of algorithm under adversarial environment by moving target defense

Kang HE; Yuefei ZHU; Long LIU; Bin LU; Bin LIU

doi:10.11959/j.issn.2096-109x.2020052

您当前的位置：

首页 >

文章列表页 >

Improve the robustness of algorithm under adversarial environment by moving target defense

Papers | 更新时间：2024-06-03

- Improve the robustness of algorithm under adversarial environment by moving target defense
- Chinese Journal of Network and Information Security Vol. 6, Issue 4, Pages: 67-76(2020)
- 作者机构：
  
  1. 信息工程大学网络空间安全学院，河南郑州450001
  2. 数学工程与先进计算国家重点实验室，河南郑州 450001
- 作者简介：
- 基金信息：
- DOI：10.11959/j.issn.2096-109x.2020052
  CLC： TP301.6
- Online First：2020-08，
  
  Published：15 August 2020
- 稿件说明：
移动端阅览
Kang HE, Yuefei ZHU, Long LIU, et al. Improve the robustness of algorithm under adversarial environment by moving target defense[J]. Chinese Journal of Network and Information Security, 2020, 6(4): 67-76.
DOI：

Kang HE, Yuefei ZHU, Long LIU, et al. Improve the robustness of algorithm under adversarial environment by moving target defense[J]. Chinese Journal of Network and Information Security, 2020, 6(4): 67-76. DOI： 10.11959/j.issn.2096-109x.2020052.

摘要

传统的机器学习模型工作在良性环境中，通常假设训练数据和测试数据是同分布的，但在恶意文档检测等领域该假设被打破。敌人通过修改测试样本对分类算法展开攻击，使精巧构造的恶意样本能够逃过机器学习算法的检测。为了提高机器学习算法的安全性，提出了基于移动目标防御技术的算法稳健性增强方法。实验证明，该方法通过在算法模型、特征选择、结果输出等阶段的动态变换，能够有效抵御攻击者对检测算法的逃逸攻击。

Abstract

Traditional machine learning models works in peace environment

assuming that training data and test data share the same distribution.However

the hypothesis does not hold in areas like malicious document detection.The enemy attacks the classification algorithm by modifying the test samples so that the well-constructed malicious samples can escape the detection by machine learning models.To improve the security of machine learning algorithms

moving target defense (MTD) based method was proposed to enhance the robustness.Experimental results show that the proposed method could effectively resist the evasion attack to detection algorithm by dynamic transformation in the stages of algorithm model

feature selection and result output.

关键词

Keywords

references

JIANG H , NAGRA J , AHAMMAD P . Sok:applying machine learning in security-a survey [J ] . arXiv preprint arXiv:1611.03186 , 2016

PITROPAKIS N , PANAOUSIS E , GIANNETSOS T , et al . A taxonomy and survey of attacks against machine learning [J ] . Computer Science Review , 2019 , 34 100199.

张东 , 张尧 , 刘刚 , 等 . 基于机器学习算法的主机恶意代码检测技术研究 [J ] . 网络与信息安全学报 , 2017 , 3 ( 7 ): 25 - 32 .

ZHANG D , ZHANG Y , LIU G , et al . Research on host malcode detection using machine learning [J ] . Chinese Journal of Network and Information Security , 2017 , 3 ( 7 ): 25 - 32 .

张骁敏 , 刘静 , 庄俊玺 , 等 . 基于权限与行为的 Android 恶意软件检测研究 [J ] . 网络与信息安全学报 , 2017 , 3 ( 3 ): 51 - 57 .

ZHANG X M , LIU J , ZHUANG J X , et al . Research on Android malware detection based on permission and behavior [J ] . Chinese Journal of Network and Information Security , 2017 , 3 ( 3 ): 51 - 57 .

SZEGEDY C , ZAREMBA W , SUTSKEVER I , et al . Intriguing properties of neural networks [J ] . arXiv preprint arXiv:1312.6199 , 2013

GOODFELLOW I J , SHLENS J , SZEGEDY C . Explaining and harnessing adversarial examples [J ] . arXiv preprint arXiv:1412.6572 , 2014

ZHANG G , YAN C , JI X , et al . Dolphinattack:inaudible voice commands [C ] // The 2017 ACM SIGSAC Conference on Computer and Communications Security . 2017 : 103 - 117 .

GROSSE K , PAPERNOT N , MANOHARAN P , et al . Adversarial perturbations against deep neural networks for malware classification [J ] . arXiv preprint arXiv:1606.04435 , 2016

CHEN S , XUE M , FAN L , et al . Automated poisoning attacks and defenses in malware detection systems:an adversarial machine learning approach [J ] . Computers ＆ Security , 2018 , 73 : 326 - 344 .

PAPERNOT N , MCDANIEL P , WU X , et al . Distillation as a defense to adversarial perturbations against deep neural networks [C ] // 2016 IEEE Symposium on Security and Privacy (SP) . 2016 : 582 - 597 .

CARLINI N , WAGNER D . Towards evaluating the robustness of neural networks [C ] // 2017 IEEE Symposium on Security and Privacy (SP) . 2017 : 39 - 57 .

蔡桂林 , 王宝生 , 王天佐 , 等 . 移动目标防御技术研究进展 [J ] . 计算机研究与发展 , 2016 , 53 ( 5 ): 968 - 987 .

CAI G L , WANG B S , WANG T Z , et al . Research and development of moving target defense technology [J ] . Journal of Computer Research and Development , 2016 , 53 ( 5 ): 968 - 987 .

EVANS D,NGUYEN-TUONG A , KNIGHT J . Effectiveness of moving target defenses [M ] // Moving Target Defense . 2011 : 29 - 48 .

JAFARIAN J H , AL-SHAER E , DUAN Q . Openflow random host mutation:transparent moving target defense using software defined networking [C ] // The First Workshop on Hot Topics in Software Defined Networks . 2012 : 127 - 132 .

SENGUPTA S , CHAKRABORTI T , KAMBHAMPATI S . MTDeep:boosting the security of deep neural nets against adversarial attacks with moving target defense [C ] // Workshops at the Thirty-Second AAAI Conference on Artificial Intelligence . 2018 .

LEI C , MA D H , ZHANG H Q . Optimal strategy selection for moving target defense based on Markov game [J ] . IEEE Access , 2017 , 5 : 156 - 169 .

ROY A , CHHABRA A , KAMHOUA C A , et al . A moving target defense against adversarial machine learning [C ] // The 4th ACM/IEEE Symposium on Edge Computing . 2019 : 383 - 388 .

李亚龙 , 陈勤 , 张旻 . 基于博弈论的移动目标最优防御策略研究 [J ] . 计算机工程与应用 , 2019 , 55 ( 19 ): 141 - 146 .

LI Y L , CHEN Q , ZHANG M . Research on optimal defense strategy of moving targets based on game theory [J ] . Computer Engineering and Applications , 2019 , 55 ( 19 ): 141 - 146 .

KANTCHELIAN A , TYGAR J D , JOSEPH A . Evasion and hardening of tree ensemble classifiers [C ] // International Conference on Machine Learning . 2016 : 2387 - 2396 .

NISSIM N , COHEN A , GLEZER C , et al . Detection of malicious PDF files and directions for enhancements:a state-of-the art survey [J ] . Computers ＆ Security , 2015 , 48 : 246 - 266 .

CHEN P Y , ZHANG H , SHARMA Y , et al . Zoo:zeroth order optimization based black-box attacks to deep neural networks without training substitute models [C ] // The 10th ACM Workshop on Artificial Intelligence and Security . 2017 : 15 - 26 .

MUÑOZ-GONZÁLEZ L , BIGGIO B , DEMONTIS A , et al . Towards poisoning of deep learning algorithms with back-gradient optimization [C ] // The 10th ACM Workshop on Artificial Intelligence and Security . 2017 : 27 - 38 .

Views

512

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Anomaly detection in DNS: an empirical study

Research progress on detection of software vulnerabilities based on code similarity

method for machine learning backdoor attacks based on defense method based on diffusion models

Personal data privacy protection method based onvertical partitioning

Malicious user detection in Momo application based on spatio-temporal behavior and static attribute analysis

Related Author

HOU Ronghao

HE Ming

KONG Kaichuan

CHEN Yong

ZUO Peng

LI Qinghao

LIU Chunling

TANG Yonghe

Related Institution

China Internet Network Information Center

School of Cyber Science and Engineering, Jinan University

Information Engineering University

School of Medical Information, Wannan Medical College

School of Physics and Electronic Information, Anhui Normal University

⁰