基于GAN的网络攻击检测研究综述

doi:10.3969/j.issn.1671-1122.2019.02.001

摘要/Abstract

摘要：

生成式对抗网络（Generative Adversarial Network,GAN）是近年来深度学习领域的一个重大突破,是一个由生成器和判别器共同构成的动态博弈模型。其“生成”和“对抗”的思想获得了广大科研工作者的青睐,满足了多个研究领域的应用需求。受该思想的启发,研究者们将GAN应用到网络安全领域,用于检测网络攻击,帮助构建智能有效的网络安全防护机制。文章介绍了GAN的基本原理、基础结构、理论发展和应用现状,着重从网络攻击样本生成、网络攻击行为检测两大方面研究了其在网络攻击检测领域的应用现状。

关键词: GAN, 生成器, 判别器, 网络攻击, 网络安全

Abstract:

Generative adversarial network (GAN) is a major breakthrough in the field of deep learning in recent years. It is a dynamic game model composed of generator and discriminator. Its ideas of “generation” and “confrontation” have won the favor of the vast number of scientific researchers and met the application needs of many research fields. Inspired by the ideas, researchers applied GAN to the field of network security to detect network attacks and help build an intelligent and effective network security protection mechanism. This paper introduces the basic principle, infrastructure, theoretical development and application status of GAN, and focuses on the application status of GAN in the field of network attack detection from two aspects of network attack sample generation and network attack behavior detection.

Key words: GAN, generator, discriminator, network attack, network security

中图分类号:

TP309

傅建明, 黎琳, 郑锐, 苏日古嘎. 基于GAN的网络攻击检测研究综述[J]. 信息网络安全, 2019, 19(2): 1-9.

Jianming FU, Lin LI, Rui ZHENG, Suriguga. Survey of Network Attack Detection Based on GAN[J]. Netinfo Security, 2019, 19(2): 1-9.

图/表 7

图1

图2

图3

表1

图4

图5

表 2

参考文献 33

[1]	GOODFELLOW I. NIPS2016 Tutorial: Generative Adversarial Networks[EB/OL].
[2]	ARJOVSKY M, BOTTOU L. Towards Principled Methods for Training Generative Adversarial Networks[EB/OL]. ,2017-1-17.
[3]	AGHAKHANI H, MACHIRY A, NILIZADEH S, et al. Detecting Deceptive Reviews Using Generative Adversarial Networks[EB/OL].,2018-5-25.
[4]	GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al.Generative Adversarial Nets[C]//NIPS. The 2014 Conference on Advances in Neural Information Processing Systems 27, December 8-13, 2014, Montreal, Canada. NY:Curran Associates, 2014. 2672-2680.
[5]	SALIMANS T, GOODFELLOW I, ZAREMBA W, et al. Improved Techniques for Training Gans[EB/OL]. ,2016-6-10.
[6]	ZHAO Junbo, MATHIEU M, LECUN Y. Energy-based Generative Adversarial Network[EB/OL]. ,2017-3-6.
[7]	ARJOVSKY M, CHINTALA S, BOTTOU L. Wasserstein Gan[EB/OL]. ,2017-12-6.
[8]	GULRAJANI I, AHMED F, ARJOVSKY M, et al. Improved Training of Wasserstein Gans[EB/OL].,2017-12-25.
[9]	LEDIG C, THEIS L, HUSZÁR F, et al. Photo-realistic Single Image Super-resolution Using a Generative Adversarial Network[EB/OL]. ,2017-5-25.
[10]	WU Huikai, ZHENG Shuai, ZHANG Junge, et al. GP-GAN: Towards Realistic High-Resolution Image Blending[EB/OL]. ,2017-3-25 .
[11]	PATHAK D, KRAHENBUHL P, DONAHUE J, et al.Context Encoders: Feature Learning by Inpainting[C]//IEEE. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, June 27-30,2016, Las Vegas,USA.NJ:IEEE,2016: 2536-2544.
[12]	MOGREN O. C-RNN-GAN: Continuous Recurrent Neural Networks with Adversarial Training[EB/OL]. ,2016-11-29.
[13]	YU Lantao, ZHANG Weinan, WANG Jun, et al.SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient[C]//AAAI. The Thirty-first AAAI Conference on Artificial Intelligence and the Twenty-ninth Innovative Applications of Artificial Intelligence Conference, February 4-9,2017,San Francisco,California,USA. CA:AAAI, 2017: 2852-2858.
[14]	LI Jiwei, MONROE W, SHI Tianlin, et al. Adversarial Learning for Neural Dialogue Generation[EB/OL]. ,2017-9-24.
[15]	ZHANG Yizhe, GAN Zhe, CARIN L. Generating Text via Adversarial Training[EB/OL]. ,2018-6-11.
[16]	MIYATO T, DAI A M, GOODFELLOW I. Adversarial Training Methods for Semi-Supervised Text Classification[EB/OL].,2017-5-6.
[17]	SUTSKEVER I, VINYALS O, LE Q V.Sequence to Sequence Learning with Neural Networks[C]//NIPS. Advances in Neural Information Processing Systems, December 8-13,2014,Montreal,Quebec,Canada.NY:Curran Associates,2016: 3104-3112.
[18]	MIKOLOV T, SUTSKEVER I, CHEN K, et al.Distributed Representations of Words and Phrases and Their Compositionality[C]//NIPS.Advances in Neural Information Processing Systems,December 5-10,2013,Lake Tahoe,Nevada,USA.NY:Curran Associates,2013: 3111-3119.
[19]	HARER J, OZDEMIR O, LAZOVICH T, et al. Learning to Repair Software Vulnerabilities with Generative Adversarial Networks[EB/OL].,2018-5-28.
[20]	YUAN Chen, QIAN Liping, ZHANG Hui, et al. Generation of Malicious Domain Training Data Based on Generative Adversarial Network[EB/OL]..
	袁辰, 钱丽萍, 张慧, 等. 基于生成对抗网络的恶意域名训练数据生成[EB/OL]..
[21]	PAN Yiming, LIN Jiajun.Generation and Verification of Malicious Network Flow Based on Generative Adversarial Networks[EB/OL].
	潘一鸣,林家骏.基于生成对抗网络的恶意网络流生成及验证[EB/OL].
[22]	HU Weiwei, TAN Ying. Generating Adversarial Malware Examples for Black-Box Attacks Based on GAN[EB/OL]. ,2017-2-20.
[23]	ZHENG Shengjun, GUO Longhua, CHEN Jian, et al.An Online Detection System for Advanced Malware Based on Virtual Execution Technology[J]. Netinfo Security, 2016,16(1) : 29-33.
	郑生军,郭龙华,陈建,等. 基于虚拟执行技术的高级恶意软件攻击在线检测系统[J]. 信息网络安全,2016,16(1) : 29-33.
[24]	KIM J Y, BU S J, CHO S B.Malware Detection Using Deep Transferred Generative Adversarial Networks[C]//ICONIP. the 24th International Conference on Neural Information Processing, November 14-18,2017,Guangzhou,China.HK:Springer,2017: 556-564.
[25]	ARNOLD A, NALLAPATI R, COHEN W W.A Comparative Study of Methods for Transductive Transfer Learning[C]//IEEE.ICDM Workshops 2007,October 28,2007,Omaha,USA.NJ:IEEE, 2007: 77-82.
[26]	MICROSOFT. Microsoft Malware Classification Challenge (BIG 2015)[EB/OL].
[27]	KIM J Y, BU S J, CHO S B.Zero-day Malware Detection Using Transferred Generative Adversarial Networks Based on Deep Autoencoders[J]. Information Sciences, 2018, 460(9): 83-102.
[28]	HEN Hongsong, WANG Gang, SONG Jianlin.Research on Anomaly Behavior Classification Algorithm of Internal Network User Based on Cloud Computing Intrusion Detection Data Set[J]. Netinfo Security, 2018 ,18(3) : 1-7.
	陈红松,王钢,宋建林. 基于云计算入侵检测数据集的内网用户异常行为分类算法研究[J]. 信息网络安全,2018,18(3):1-7.
[29]	YIN Chuanlong, ZHU Yuefei, LIU Shengli, et al.An Enhancing Framework for Botnet Detection Using Generative Adversarial Networks[C]//IEEE.2018 International Conference on Artificial Intelligence and Big Data (ICAIBD),May 26-28,2018,Chengdu,China.NJ:IEEE, 2018:228-234.
[30]	SHIRAVI A, SHIRAVI H, TAVALLAEE M, et al.Toward Developing a Systematic Approach to Generate Benchmark Datasets for Intrusion Detection[J]. Computers & Security, 2012, 31(3): 357-374.
[31]	FIORE U, DE S A, PERLA F, et al. Using Generative Adversarial Networks for Improving Classification Effectiveness in Credit Card Fraud Detection[EB/OL]. ,2017-12-30.
[32]	DAL P A, CAELEN O, JOHNSON R A, et al.Calibrating Probability with Undersampling for Unbalanced Classification[C]//IEEE. 2015 IEEE Symposium Series on Computational Intelligence, December 7-10,2015,Cape Town,South Africa.NJ:IEEE,2015: 159-166.
[33]	ZHENG Yujun, ZHOU Xiaohan, SHENG Weiguo, et al.Generative Adversarial Network-based Telecom Fraud Detection at the Receiving Bank[J]. Neural Networks, 2018, 102(6): 78-86.

项目模型	特征选择	训练数据有无标签	检测结果输出	应用领域
t-GAN	无需人工选取	无	二分类（正常/恶意）	无监督学习
t-DCGAN	无需人工选取	无	二分类（正常/恶意）	无监督学习
Bot-GAN	人工选取	无	三分类（正常/ 异常/虚假）	无监督学习
CF-GAN	无需人工选取	有	二分类（合法/违法）	有监督学习
GAN-DAE	无需人工选取	有	三分类（正常/ 欺诈/未知）	有监督学习