Seeing Traffic Paths: Encrypted Traffic Classification With Path Signature Features
Визуализация маршрутов трафика: классификация зашифрованного трафика с использованием признаков сигнатур маршрута
2022-01-01
SCID: 54.1/hvv75m65
Discuss with AI
QUIC and HTTPS trafficencrypted traffic classificationmultiscale path signaturepath signature featuressession packet length sequence
Figures from the paper
Abstract (AI)
Although many network traffic protection methods have been developed to protect user privacy, encrypted traffic can still reveal sensitive user information with sophisticated analysis. In this paper, we propose ETC-PS, a novel encrypted traffic classification method with path signature. We first construct the traffic path with a session packet length sequence to represent the interactions between the client and the server. Then, path transformations are conducted to exhibit its structure and obtain different information. A multiscale path signature is finally computed as a kind of distinctive feature to train the traditional machine learning classifier, which achieves highly robust accuracy and low training overhead. Six publicly available datasets with different traffic types of HTTPS/1, HTTPS/2, QUIC, VPN, non-VPN, Tor, and non-Tor are used to conduct closed-world and open-world evaluations to verify the effectiveness of ETC-PS. The experimental results demonstrate that ETC-PS is superior to the state-of-the-art methods in terms of accuracy, f1 score, time complexity, and stability.
Key Findings
1
ETC-PS classifies encrypted traffic by representing client–server interactions as session packet-length paths.
2
ETC-PS outperforms state-of-the-art methods in accuracy, F1 score, time complexity, and stability.
3
Evaluations across six public datasets covering HTTPS/1, HTTPS/2, QUIC, VPN, non-VPN, Tor, and non-Tor traffic include both closed-world and open-world settings.
4
Path transformations expose traffic-path structure, while multiscale path signatures provide distinctive features for traditional machine-learning classifiers.
5
The method achieves highly robust classification accuracy with low training overhead.
Research Object
encrypted network traffic from diverse protocols and network environments
Research Subject
traffic classification performance and robustness based on path-signature features derived from client–server packet-length sequences
Publication Details
Publication Date
2022-01-01
Journal
Publisher
ISSN
Cited by
111
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai5
End-to-end encrypted traffic classification with one-dimensional convolution neural networks2017
FS-Net: A Flow Sequence Network For Encrypted Traffic Classification2019
MaMPF: Encrypted Traffic Classification Based on Multi-Attribute Markov Probability Fingerprints2018
Learning to Classify: A Flow-Based Relation Network for Encrypted Traffic Classification2020
HTTPS traffic analysis and client identification using passive SSL/TLS fingerprinting2016