FlowAtom improves multi-label website identification from encrypted traffic
FlowAtom: Atom-Based Evidence Aggregation for Multi-Label Website Fingerprinting
Machine LearningNetworking and Internet Architecture
Summary
It’s hard to figure out which websites someone visited when their internet traffic is encrypted and mixed together. The authors propose FlowAtom, a way to combine small parts of internet data, called Atoms, from different pieces of traffic to recognize sets of visited websites better. They train FlowAtom without knowing website labels ahead of time and show it works well on different encrypted traffic types. Their approach beats other methods in tests where multiple monitored websites appear together.
What this means in practice
- •For network security teams: Detect multiple monitored websites visited simultaneously from encrypted traffic for enhanced monitoring in sensitive environments.
- •For internet privacy tool developers: Improve privacy tools’ understanding of fingerprinting risks when multiple websites are accessed at once over encrypted connections.
Authors
Chongru Fan, Wentao Huang, Wei Wang, Zhenquan Ding, Jinqiao Shi, Wei Cai, Zhiyu Hao
Abstract
Identifying the set of monitored websites in mixed encrypted traffic is challenging because an individual flow often provides only partial evidence of website identity. To address this challenge, we propose FlowAtom, which constructs shared prototypes, called Atoms, from flow representations without website labels. Specifically, FlowAtom pretrains a flow encoder on external unlabeled traffic and aggregates Atom responses across flows within each observation window into a fixed-dimensional, permutation-invariant representation for monitored website-set prediction. Across Direct HTTPS, Trojan, and VMess, FlowAtom achieves micro-F1 scores of 97.82%, 94.43%, and 93.92% in closed-world evaluation, respectively, and consistently outperforms the evaluated baselines in open-world evaluation on windows containing monitored visits. The code is available at https://github.com/aimafan123/FlowAtom.