面向网络入侵检测数据的对抗样本生成方法

doi:10.6040/j.issn.1671-9352.4.2020.152

摘要/Abstract

摘要：

选用Deepfool以及JSMA(jacobian-based saliency map attack)算法，在攻击特征中加入不影响攻击特性的定向扰动，通过白盒攻击生成对抗样本。通过实现扰乱检测模型的判断，从而躲过特征检测，为入侵检测模型提升自身鲁棒性提供了更为丰富的训练样本。

关键词: 网络入侵检测, 神经网络, 对抗样本, KDD Cup99

Abstract:

This paper proposes to add directional perturbations having no impact on results to attack characteristics with Deepfool and JSMA algorithms. Adversarial samples are generated by white-box attacks so that they can interfere with the judgements of models to bypass feature detection. Our work provides intrusion detection models with more training samples. As a result, the robustness of intrusion detection models is improved.

Key words: network intrusion detection, neural network, adversarial example, KDD Cup99

中图分类号:

TP399

解滨,李清扬,董新玉. 面向网络入侵检测数据的对抗样本生成方法[J]. 《山东大学学报(理学版)》, 2021, 56(3): 28-36.

Bin XIE,Qing-yang LI,Xin-yu DONG. Adversarial examples generation method for network intrusion detection data[J]. JOURNAL OF SHANDONG UNIVERSITY(NATURAL SCIENCE), 2021, 56(3): 28-36.

图/表 10

图1

图2

图3

图4

图5

表1

表2

表3

图6

图7

参考文献 21

1	刘浩然, 丁攀, 郭长江, 等. 基于贝叶斯算法的中文垃圾邮件过滤系统研究[J]. 通信学报, 2018, 39 (12): 151- 159.
	LIU Haoran , DING Pan , GUO Changjiang , et al. Study on Chinese spam filtering system based on Bayes algorithm[J]. Journal on Communications, 2018, 39 (12): 151- 159.
2	彭成维, 云晓春, 张永铮, 等. 一种基于域名请求伴随关系的恶意域名检测方法[J]. 计算机研究与发展, 2019, 56 (6): 1263- 1274.
	PENG Chengwei , YUN Xiaochun , ZHANG Yongzheng , et al. Detecting malicious domains using co-occurrence relation between DNS query[J]. Journal of Computer Research and Development, 2019, 56 (6): 1263- 1274.
3	刘金平, 周嘉铭, 刘先锋, 等. 基于聚类簇结构特性的自适应综合采样法在入侵检测中的应用[J/OL]. 控制与决策, (2020-03-31)[2020-04-2]http://kns.cnki.net/kcms/detail/21.1124.TP.20200330.1533.033.html.
	LIU Jinping, ZHOU Jiaming, LIU Xianfeng, et al. Toward intrusion detection via cluster structure-based adaptive synthetic sampling approach[J/OL]. Control and Decision, (2020-03-31)[2020-03-28]http://kns.cnki.net/kcms/detail/21.1124.TP.20200330.1533.033.html.
4	江颉, 高甲, 陈铁明. 基于AE-BNDNN模型的入侵检测方法[J]. 小型微型计算机系统, 2019, 40 (8): 1713- 1717. doi: 10.3969/j.issn.1000-1220.2019.08.025
	JIANG Jie , GAO Jia , CHEN Tieming . Network intrusion detection method based on AE-BNDNN model[J]. Journal of Chinese Computer Systems, 2019, 40 (8): 1713- 1717. doi: 10.3969/j.issn.1000-1220.2019.08.025
5	SZEGEDY C, ZAREMBA W, SUTSKEVER I, et al. Intriguing properties of neural networks[EB/OL]. (2014-02-19)[2020-03-28]. https: //arxiv.org/abs/1312.6199
6	潘文雯, 王新宇, 宋明黎, 等. 对抗样本生成技术综述[J]. 软件学报, 2020, 31 (1): 67- 81.
	PAN Wenwen , WANG Xinyu , SONG Mingli , et al. Survey on generating adversarial examples[J]. Journal of Software, 2020, 31 (1): 67- 81.
7	陈岳峰, 毛潇锋, 李裕宏, 等. AI安全: 对抗样本技术综述与应用[J]. 信息安全研究, 2019, 5 (11): 1000- 1007. doi: 10.3969/j.issn.2096-1057.2019.11.009
	CHEN Yuefeng , MAO Xiaofeng , LI Yuhong , et al. AI security: research and application on adversarial example[J]. Journal of Information Security Research, 2019, 5 (11): 1000- 1007. doi: 10.3969/j.issn.2096-1057.2019.11.009
8	易平, 王科迪, 黄程. 人工智能对抗攻击研究综述[J]. 上海交通大学学报, 2018, 52 (10): 1298- 1306.
	YI Ping , WANG Kedi , HUANG Cheng . Adversarial attacks in artificial intelligence: a survey[J]. Journal of Shanghai Jiaotong University, 2018, 52 (10): 1298- 1306.
9	张蕾, 崔勇, 刘静, 等. 机器学习在网络空间安全研究中的应用[J]. 计算机学报, 2018, 41 (9): 1943- 1975.
	ZHANG Lei , CUI Yong , LIU Jing , et al. Application of machine learning in cyberspace security research[J]. Chinese Journal of Computers, 2018, 41 (9): 1943- 1975.
10	王晓程, 刘恩德, 谢小权. 攻击分类研究与分布式网络入侵检测系统[J]. 计算机研究与发展, 2001, 38 (6): 727- 734.
	WANG Xiaocheng , LIU Ende , XIE Xiaoquan . Attack classification research and a distributed network intrusion detection system[J]. Journal of Computer Research and Development, 2001, 38 (6): 727- 734.
11	杨印根, 王忠洋. 基于深度神经网络的入侵检测技术[J]. 网络安全技术与应用, 2019, (4): 37- 41.
	YANG Yingen , WANG Zhongyang . Intrusion detection technology based on deep neural network[J]. Network Security Technology & Application, 2019, (4): 37- 41.
12	SEYED-MOHSEN M D, ALHUSSEIN F. DeepFool: a simple and accurate method to fool deep neural networks[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas: IEEE Computer Society, 2016: 2574-2582.
13	PAPERNOT N, MCDANIEL P, JHA S, et al. The limitations of deep learning in adversarial settings[C]//IEEE European Symposium on Security and Privacy (EuroS&P). Saarbrucken: IEEE, 2016: 372-387.
14	ZONG B, SONG Q, MIN M R, et al. Deep autoencoding Gaussian mixture model for unsupervised anomaly detection[EB/OL]. (2018-03-01)[2020-03-28]https://openreview.net/pdf?id=BJJLHbb0-.
15	李小剑, 谢晓尧. 基于支持向量机与k近邻相结合的网络入侵检测研究[J]. 贵州师范大学学报(自然科学版), 2015, 33 (3): 86- 91.
	LI Xiaojian , XIE Xiaoyao . Research on network intrusion detection based on support vector machine combined with k nearest neighbor method[J]. Journal of Guizhou Normal University(Natural Sciences), 2015, 33 (3): 86- 91.
16	GOODFELLOW I, SHLENS J, SZEGEDY C. Explaining and harnessing adversarial examples[C]//3rd International Conference on Learning Representations(ICLR). San Diego: Computer Science, 2015.
17	BRENDEL W, RAUBER J, BETHGE M. Decision-based adversarial attacks: reliable attacks against black-box machine learning models[EB/OL]. (2018-02-16)[2020-03-28]. https://arxiv.org/abs/1712.04248.
18	PAPERNOT N, MCDANIEL P, GOODFELLOW I. Transferability in machine learning: from phenomena to black-box attacks using adversarial samples[EB/OL]. (2016-5-24)[2020-03-28]. https://arxiv.org/abs/1605.07277.
19	张玉清, 董颖, 柳彩云, 等. 深度学习应用于网络空间安全的现状、趋势与展望[J]. 计算机研究与发展, 2018, 55 (6): 1117- 1142.
	ZHANG Yuqing , DONG Ying , LIU Caiyun , et al. Situation, trends and prospects of deep learning applied to cyberspace security[J]. Journal of Computer Research and Development, 2018, 55 (6): 1117- 1142.
20	DIEDERIK P K, JIMMY L B. Adam: a method for stochastic optimization[C]//3rd International Conference on Learning Representations(ICLR). San Diego: Computer Science, 2015.
21	聂凯, 周清雷, 朱维军, 等. 基于时序逻辑的3种网络攻击建模[J]. 计算机科学, 2018, 45 (2): 209- 214.
	NIE Kai , ZHOU Qinglei , ZHU Weijun , et al. Modeling for three kinds of network attacks based on temporal logic[J]. Computer Science, 2018, 45 (2): 209- 214.

相关文章 15

[1]	银温社,贺建峰. 基于深度学习的眼底图像出血点检测方法[J]. 《山东大学学报(理学版)》, 2020, 55(9): 62-71.
[2]	温晓,刘琪,高振,曾维新,吕咸青. 局部非侵入式约化基模型在瑞利-泰勒不稳定中的应用[J]. 《山东大学学报(理学版)》, 2020, 55(2): 109-117.
[3]	王文卿,撖奥洋,于立涛,张智晟. 自编码器与PSOA-CNN结合的短期负荷预测模型[J]. 《山东大学学报(理学版)》, 2019, 54(7): 50-56.
[4]	戴丽华,惠远先. 时标上具有联接项时滞的分流抑制细胞神经网络的概自守解[J]. 《山东大学学报(理学版)》, 2019, 54(10): 97-108.
[5]	刘飚,路哲,黄雨薇,焦萌,李泉其,薛瑞. 神经网络结构在功耗分析中的性能对比[J]. 《山东大学学报(理学版)》, 2019, 54(1): 60-66.
[6]	肖炜茗,王贵君. 基于Bernstein多项式的SISO三层前向神经网络的设计与逼近[J]. 山东大学学报（理学版）, 2018, 53(9): 55-61.
[7]	张芳芳,曹兴超. 基于字面和语义相关性匹配的智能篇章排序[J]. 山东大学学报（理学版）, 2018, 53(3): 46-53.
[8]	刘明明,张敏情,刘佳,高培贤. 一种基于浅层卷积神经网络的隐写分析方法[J]. 山东大学学报（理学版）, 2018, 53(3): 63-70.
[9]	李翠平,高兴宝. 求解具有约束的l₁-范数问题的神经网络模型[J]. 《山东大学学报(理学版)》, 2018, 53(12): 90-98.
[10]	秦静,林鸿飞,徐博. 基于示例语义的音乐检索模型[J]. 山东大学学报（理学版）, 2017, 52(6): 40-48.
[11]	王长弘,王林山. 基于忆阻器的S-分布时滞随机神经网络的均方指数稳定性[J]. 山东大学学报（理学版）, 2016, 51(5): 130-135.
[12]	甄艳, 王林山. S-分布时滞随机广义细胞神经网络的均方指数稳定性分析[J]. 山东大学学报（理学版）, 2014, 49(12): 60-65.
[13]	杨阳, 刘龙飞, 魏现辉, 林鸿飞. 基于词向量的情感新词发现方法[J]. 山东大学学报（理学版）, 2014, 49(11): 51-58.
[14]	刘铭, 昝红英, 原慧斌. 基于SVM与RNN的文本情感关键句判定与抽取[J]. 山东大学学报（理学版）, 2014, 49(11): 68-73.
[15]	堵锡华,史小琴,冯长君,李亮. 基于野韭菜挥发性成分的色谱保留指数神经网络预测[J]. 山东大学学报（理学版）, 2014, 49(1): 50-53.

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed

训练的模型	扰动特征(变化前→变化后)	范数	DAGMM预测值	KNN预测值
1	num_root (0→3)	L₀=1，L₂=23.67	改变判断(7.085→19.374)	不变
2	hot (1→0)	L₀=3，L₂=2.26	改变判断(7.085→12.611)	不变
	num_root (0→1)
	dst_host_count (70→7)
3	duration (0→360)	L₀=6，L₂=1.55	不改变判(7.085→7.043)	不变
	dst_bytes (7300→26604)
	srv_diff_host_rate (0→0.05)
	dst_host_count (70→29)
	srv_diff_host_rate (70→188)
	dst_host_rerror_rate(0.2→0.13)

训练的模型	扰动特征(变化前→变化后)	范数	DAGMM预测值	KNN预测值
1	num_root (0->3)	L₀=1，L₂=23.67	改变判断(7.085->19.374)	不变
2	num_file_creations (0→2)	L₀=2，L₂=23.67	改变判断(7.085→7.671)	不变
	dst_host_diff_srv_rate (0→0.1)
3	num_root(0→15)	L₀=1，L₂=22.88	改变判断(7.085→20.076)	不变

对抗算法	平均迭代次数	平均更改特征数(直接生成不调整)	平均对抗成功率/%
Deepfool	3	18	94.4
JSMA	4	12	97.9