基于去噪动态窗循环神经网络的混淆网页流量识别方法
网络出版日期: 2025-01-25
基金资助
国家重点研发计划(2021QY0700);国家自然科学基金(U21B2003,62072250)
版权
Obfuscated web traffic recognition method based on denoising dynamic window recurrent neural network
Online published: 2025-01-25
Copyright
对网页流量进行混淆是当前很多非法网络通联突破监管的常用手段,有效识别网页流量对网络安全空间的防护至关重要。在网络流量识别领域,深度学习技术已成为突破传统方法局限性、提高识别准确率的关键。提出一种基于去噪动态窗循环神经网络(Denoising Dynamic Window Recurrent Neural Network,DDWRNN)的混淆加密网页流量识别方法,旨在高效地识别经过混淆处理的加密网页流量。首先基于去噪自编码器(Denoising Autoencoder,DAE)的降噪和特征提取能力有效提取具有较强鲁棒性的流量深层次时空特征,其次通过设计的动态窗循环神经网络的多尺度动态时间窗口捕获流量在不同时间尺度上的动态特征。实验结果表明,混淆加密网页流量识别的平均准确率达到95.7%,与其他机器学习、深度学习方法相比,提出的混淆加密网页流量识别方法不仅在识别准确率上有所提升,而且在处理复杂的流量模式、适应动态变化的网络环境方面也展现出较强的能力。
饶晨 , 黄佳男 , 刘伟伟 . 基于去噪动态窗循环神经网络的混淆网页流量识别方法[J]. 网络空间安全科学学报, 2024 , 2(6) : 98 -109 . DOI: 10.20172/j.issn.2097-3136.240607
Obfuscation of the web traffic is a common tactic used by many illicit online communications to circumvent monitoring, making the effective web traffic recognition crucial for cyberspace protection. In the field of network traffic recognition, the deep learning technology has become a key to break through the limitations of traditional methods and to enhance the recognition accuracy. A method for the obfuscated encrypted web traffic recognition based on the denoising dynamic window recurrent neural network (DDWRNN) was proposed, aiming to efficiently identify the obfuscated encrypted web traffic. Firstly, leveraging the denoising and feature extraction capabilities of denoising autoencoders (DAE), the robust deep spatiotemporal features were effectively extracted from the traffic data. Subsequently, the designed dynamic window recurrent neural network captured the dynamic features of traffic across various time scales using its multi-scale dynamic temporal windows. Experimental results showed that the average recognition accuracy of the proposed method reached 95.7% for the obfuscated encrypted web page traffic. Compared to other machine learning and deep learning approaches, the proposed method not only showed an improvement of the recognition accuracy but also demonstrated strong capabilities to handle the complex traffic patterns and to adapt to the dynamically changing network environments.
表 1 服务器配置与开发环境Table 1 Server configuration and development environment |
| 项目 | 配置信息 |
| 操作系统 | Windows 10 64位 |
| CPU | Intel(R) Xeon(R) CPU E5- |
| GPU | NVIDIA GeForce GTX 1080Ti/PCle/SSE2 |
| 内存大小 | 64 GB |
| 磁盘空间 | 1 TB |
| 开发环境 | Python-3.7.9/Tensorflow-1.13.1/Numpy-1.20.3/ Pandas-1.3.5/Scikit-learn-1.0.2 |
表 2 实验数据集信息Table 2 Experimental dataset information |
| 数据集 | 网页名称 | 样本流数量 | 总计 |
| 训练集 | |||
| apple | |||
| douyin | |||
| 测试集 | |||
| apple | 927 | ||
| douyin | 1287 | ||
| 验证集 | |||
| apple | 928 | ||
| douyin |
图 4 不同时间窗口尺度下DDWRNN模型识别准确率变化Fig.4 Variation in recognition accuracy of the DDWRNN model under different time window scales |
表 3 DDWRNN模型相关参数Table 3 DDWRNN model related parameters |
| 参数项目 | 参数数值 |
| AE层数 | 3 |
| 输入维度 (Layer 1) | 6 |
| 输入维度 (Layer 2) | 128 |
| 输入序列时间步数 | 180 |
| 隐藏状态数目(Layer 1 & Layer 2) | 128 |
| 隐藏状态窗口尺度向量 | |
| Input_Dropout (Layer 1) | 0.1 |
| Input_Dropout (Layer 2) | 0.25 |
| batch_size | 16 |
| epoch | 20 |
| 初始学习速率 | 0.001 |
| 优化器 | Adam |
| 损失函数 | Cross_Entropy |
| 1 |
CNNIC. 第52次《中国互联网络发展状况统计报告》[EB/OL]. (2023-8-28)[2023-10-20]. http://www.cnnic.net.cn/n4/2023/0828/c199-10830.html.
CNNIC. The 52nd statistical report on Chinese Internet development [EB/OL]. (2023-8-28)[2023-10-20]. http://www.cnnic.net.cn/n4/2023/0828/c199-10830.html.
|
| 2 |
PAPADOGIANNAKI E, IOANNIDIS S. A survey on encrypted network traffic analysis applications, techniques, and countermeasures[J]. ACM Computing Surveys, 2021, 54 (6): 1- 35.
|
| 3 |
ACETO G, CIUONZO D, MONTIERI A, et al. Mobile encrypted traffic classification using deep learning: experimental evaluation, lessons learned, and challenges[J]. IEEE transactions on network and service management, 2019, 16 (2): 445- 458.
|
| 4 |
SHEN M, LIU Y, ZHU L, et al. Optimizing feature selection for efficient encrypted traffic classification: a systematic approach[J]. IEEE Network, 2020, 34 (4): 20- 27.
|
| 5 |
MAIOLINI G,BAIOCCHI A,IACOVAZZI A,et al. Real time identification of SSH encrypted application flows by using cluster analysis techniques[C]//NETWORKING 2009:8th International IFIP-TC 6 Networking Conference. Aachen,Germany:May 11-15,2009. Proceedings 8. Springer Berlin Heidelberg,2009:182-194.
|
| 6 |
ZENG X, CHEN X, SHAO G, et al. Flow context and host behavior based shadowsocks’s traffic identification[J]. IEEE Access, 2019, 7, 41017- 41032.
|
| 7 |
DENG Z,LIU Z,CHEN Z,et al. The random forest based detection of shadowsock's traffic[C]//2017 9th International Conference on Intelligent Human-Machine Systems and Cybernetics (IHMSC). IEEE,2017,2:75-78.
|
| 8 |
SHEN M, YE K, LIU X, et al. Machine learning-powered encrypted network traffic analysis: a comprehensive survey[J]. IEEE Communications Surveys & Tutorials, 2022, 25 (1): 791- 824.
|
| 9 |
ZANG M,ZHENG C,DITTMANN L,et al. Towards continuous threat defense:In-Network traffic analysis for IoT gateways[J]. IEEE Internet of Things Journal,2023.
|
| 10 |
HAFEEZ I,ANTIKAINEN M,TARKOMA S. Protecting IoT-environments against traffic analysis attacks with traffic morphing[C]//2019 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops). IEEE,2019:196-201.
|
| 11 |
JANSEN R,JOHNSON A. Safely measuring tor[C]//Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security,2016:1553-1567.
|
| 12 |
MOORE W B,TAN H,SHERR M,et al. Multi-class traffic morphing for encrypted voip communication[C]//Financial Cryptography and Data Security:19th International Conference,FC 2015. San Juan,Puerto Rico:January 26-30,2015,Revised Selected Papers 19. Springer Berlin Heidelberg,2015:65-85.
|
| 13 |
VERMA G,CIFTCIOGLU E,SHEATSLEY R,et al. Network traffic obfuscation:an adversarial machine learning approach[C]//MILCOM 2018-2018 IEEE military communications conference (MILCOM). IEEE,2018:1-6.
|
| 14 |
LIU L, YU H, YU S, et al. Network traffic obfuscation against traffic classification[J]. Security and Communication Networks, 2022, 2022 (1): 3104392.
|
| 15 |
GRANADOS A,MIAH M S,ORTIZ A,et al. A realistic approach for network traffic obfuscation using adversarial machine learning[C]//Decision and Game Theory for Security:11th International Conference,GameSec 2020. College Park,MD,USA:October 28–30,2020,Proceedings 11. Springer International Publishing,2020:45-57.
|
| 16 |
DUSI M, CROTTI M, GRINGOLI F, et al. Tunnel hunter: detecting application-layer tunnels with statistical fingerprinting[J]. Computer Networks, 2009, 53 (1): 81- 97.
|
| 17 |
JUAREZ M,IMANI M,PERRY M,et al. Toward an efficient website fingerprinting defense[C]//Computer Security–ESORICS 2016:21st European Symposium on Research in Computer Security. Heraklion,Greece:September 26-30,2016,Proceedings,Part 1 21. Springer International Publishing,2016:27-46.
|
| 18 |
WRIGHT C V,COULL S E,MONROSE F. Traffic morphing:an efficient defense against statistical traffic analysis[C]//NDSS. 2009:9.
|
| 19 |
LI J,ZHOU L,LI H,et al. Dynamic traffic feature camouflaging via generative adversarial networks[C]//2019 IEEE Conference on Communications and Network Security (CNS). IEEE,2019:268-276.
|
| 20 |
HUANG G,MA C,DING M,et al. Efficient and low overhead website fingerprinting attacks and defenses based on TCP/IP traffic[C]//Proceedings of the ACM Web Conference.2023:1991-1999.
|
| 21 |
DE LUCIA M J,COTTON C. Detection of encrypted malicious network traffic using machine learning[C]//MILCOM 2019-2019 IEEE Military Communications Conference (MILCOM). IEEE,2019:1-6.
|
| 22 |
王曦锐, 芦天亮, 张建岭, 等. 基于加权Stacking集成学习的Tor匿名流量识别方法[J]. 信息网络安全, 2021, 21 (12): 118- 125.
WANG X R, LU T L, ZHANG J L, et al. Tor anonymous traffic identification method based on weighted stacking ensemble learning[J]. Netinfo Security, 2021, 21 (12): 118- 125.
|
| 23 |
SHEN M,JI K,GAO Z,et al. Subverting website fingerprinting defenses with robust traffic representation[C]//32nd USENIX Security Symposium (USENIX Security 23). 2023:607-624.
|
| 24 |
TANG J,YANG L,LIU S,et al. Caps-lstm:a novel hierarchical encrypted VPN network traffic identification using capsnet and LSTM[C]//International Conference on Science of Cyber Security. Cham:Springer International Publishing,2021:139-153.
|
| 25 |
蒋彤彤, 尹魏昕, 蔡冰, 等. 基于层次时空特征与多头注意力的恶意加密流量识别[J]. 计算机工程, 2021, 47 (7): 101- 108.
JIANG T T, YIN W X, CAI B, et a1. Encrypted malicious traffic identification based on hierarchical spatiotemporal feature and multi-head attention[J]. Computer Engineering, 2021, 47 (7): 101- 108.
|
| 26 |
HOLLAND J,SCHMITT P,FEAMSTER N,et al. New directions in automated traffic analysis[C]//Proceedings of the 2021 ACM SIGSAC Conference on Computer and Communications Security. 2021:3366-3383.
|
| 27 |
NAYAK S C, MISRA B B, BEHERA H S. Impact of data normalization on stock index forecasting[J]. International Journal of Computer Information Systems and Industrial Management Applications, 2014, 6, 257- 269.
|
| 28 |
SINGH D, SINGH B. Investigating the impact of data normalization on classification performance[J]. Applied Soft Computing, 2020, 97, 105524.
|
| 29 |
ABE K, GOTO S. Fingerprinting attack on Tor anonymity using deep learning[J]. Proceedings of the Asia-Pacific Advanced Network, 2016, 42, 15- 20.
|
| 30 |
HERMANS M,SCHRAUWEN B. Training and analysing deep recurrent neural networks[J]. Advances in Neural Information Processing Systems,2013,26.
|
| 31 |
KOUTNIK J,GREFF K,GOMEZ F,et al. A clockwork RNN[C]//International Conference on Machine Learning. PMLR,2014:1863-1871.
|
/
| 〈 |
|
〉 |