CipherSight: Robust Website Fingerprinting via Record-Resource Semantic Supervision under Distribution Shifts

Runhan Song; Qiqi Liu; Chuanzhou Pan; Zhenquan Ding; Youquan Xian; Chongru Fan; Lei Cui; Wei Wang

cs.CRAug 14, 2026

CipherSight: Robust Website Fingerprinting via Record-Resource Semantic Supervision under Distribution Shifts

Authors: Runhan Song, Qiqi Liu, Chuanzhou Pan, Zhenquan Ding, Youquan Xian, Chongru Fan, Lei Cui, Wei Wang,

Organizations: Faculty of Computing, Harbin Institute of Technology · Zhongguancun Laboratory · University of Chinese Academy of Sciences · School of Cyberspace Security, Beijing University of Posts and Telecommunications

Abstract

HTTPS website fingerprinting (WF) aims to identify visited websites from metadata observable in encrypted traffic. However, real-world deployments introduce a significant out-of-distribution (OOD) problem caused by temporal and geographic changes, while previously unseen websites are common in open-world scenarios. Existing methods primarily learn from raw TCP packet sequences and struggle to capture stable and generalizable website representations, resulting in performance degradation under practical conditions. We propose CipherSight, a TLS-record-based hierarchical framework for robust HTTPS WF. Unlike existing approaches that rely on TCP packet sequences and are sensitive to transport-layer artifacts, CipherSight learns website representations from TLS records by jointly encoding multiple record-level attributes. It introduces a hierarchical architecture that captures both intra-flow dependencies among TLS records and inter-flow interactions across concurrent flows, enabling the model to exploit structural patterns in HTTPS traffic. Besides, to learn robust representations, CipherSight employs a masked record modeling (MRM) task to capture contextual traffic semantics and leverages fine-grained record-resource annotations as privileged supervision through structure-aware objectives and semantic distillation. Experiments show that CipherSight achieves 95.41% accuracy across more than 2,000 website classes in the closed-world setting and maintains over 90% accuracy under both temporal and geographic drift, consistently outperforming all evaluated baselines.

Figures & tables

Figure 1: Challenges and motivation for CipherSight. Robust HTTPS WF requires reduced sensitivity to transport variability, explicit modeling of dependencies within and across flows, and supervision that connects encrypted records with web-resource structure.

Figure 2: Overview of CipherSight . During representation construction, observable TLS records are grouped by flow and tokenized. Block-diagonal intra-flow attention (L1) encodes record tokens within each flow, while full inter-flow attention (L2) aggregates flow summaries for website classification. Training-time privileged structure provides record-resource supervision (P1) and semantic teacher distillation (P2).

Figure 3: Fine-tuning architecture of CipherSight. The LoRA-adapted ciphertext student receives resource-structure supervision (P1), while a privileged semantic teacher (P2) transfers resource-level knowledge through feature and logit distillation, $L_{\mathrm{feat}}$ and $L_{\mathrm{logit}}$ . The student and teacher are supervised by $L_{\mathrm{cls}}^{S}$ and $L_{\mathrm{cls}}^{T}$ , respectively. Only the student is retained for inference.

Model	Origin	Accuracy	Precision	Recall	Macro-F1	Top-5
VarCNN	Tor	92.16%	92.63%	92.02%	91.91%	97.27%
ARES	Tor	91.22%	91.65%	91.07%	90.85%	97.11%
RF	Tor	91.09%	91.27%	90.84%	90.63%	97.16%
DF	Tor	89.85%	90.29%	89.56%	89.36%	96.61%
TikTok	Tor	89.32%	89.69%	89.04%	88.77%	96.50%
TF	Tor	86.94%	87.37%	86.62%	86.41%	–

Table 1: Closed-world performance comparison on 2,008 website classes. CipherSight results are reported as the mean

\pm

standard deviation over five seeds.

Model	Accuracy (%)	Macro-F1 (%)
VarCNN	67.81 (-24.35)	67.30 (-24.61)
ARES	72.84 (-18.38)	71.92 (-18.93)
RF	66.07 (-25.02)	65.82 (-24.81)
DF	67.92 (-21.93)	66.61 (-22.75)
TikTok	66.03 (-23.29)	64.74 (-24.03)
TF	67.75 (-19.19)	66.49 (-19.92)

Table 2: Temporal-drift performance from us0304 to us0320 with a 16-day collection gap. Values in parentheses denote descriptive percentage-point differences from each model’s closed-world result on us0304 .

	Singapore			France
Model	Accuracy (%)	Macro-F1 (%)	Top-5 (%)	Accuracy (%)	Macro-F1 (%)	Top-5 (%)
VarCNN	67.53 (-24.98)	66.20 (-26.24)	83.07 (-14.30)	73.12 (-19.39)	70.70 (-21.74)	85.05 (-12.32)
ARES	79.53 (-13.16)	78.15 (-14.48)	89.97 (- 0 7.23)	81.08 (-11.61)	79.11 (-13.52)	90.58 (- 0 6.62)
RF	57.69 (-35.12)	57.44 (-35.19)	73.25 (-24.10)	82.07 (-10.74)	80.19 (-12.44)	91.03 (- 0 6.32)
DF	74.95 (-17.39)	72.83 (-19.26)	87.20 (-10.10)	76.48 (-15.86)	73.24 (-18.85)	86.41 (-10.89)
TikTok	70.00 (-21.98)	67.80 (-23.94)	84.26 (-13.14)	74.51 (-17.47)	71.24 (-20.50)	85.08 (-12.32)

Table 3: Geographic-drift performance from us0409 to Singapore and France. Values in parentheses denote descriptive percentage-point differences from each model’s us0409 source-domain holdout. CipherSight values are means over five seeds.

Test	No-Priv (%)	P1 Pre. (%)	Full Priv. (%)
us0320	92.69 / 91.73	92.99 / 92.07	92.99 / 92.07
us0409	90.46 / 89.02	90.93 / 89.53	91.11 / 89.73
fr0409	89.04 / 87.37	90.15 / 88.73	90.59 / 89.19
sg0409	89.73 / 88.42	90.59 / 89.46	91.09 / 89.99

Table 4: Ablation of privileged supervision under temporal and geographic drift. No-Priv. excludes P1 and P2, P1 Pre. applies P1 only during pretraining, and Full Priv. applies P1 in both stages and introduces P2 during LoRA fine-tuning. Cells report mean accuracy/macro-F1 (%) over five seeds.

Model	AUROC (%)	AUPR (%)	OSCR (%)
VarCNN	89.56	60.50	86.22
ARES	89.93	67.27	86.05
RF	93.04	72.03	87.61
DF	87.06	63.46	83.72
TikTok	86.21	61.49	82.77
TF	82.75	26.95	77.15

Table 5: Open-world evaluation results using maximum softmax probability (MSP) as the confidence score.

Encrypted Network Traffic Analysis

Encrypted network traffic analysis infers applications, threats, and network conditions from traffic patterns without decrypting payloads.

See all 29 papers

ResAware: Cross-Environment Website Fingerprinting via Resource-Privileged Distillation
Jun 16, 2026Chongru Fan, Wei Wang, Wentao Huang +5Encrypted Network Traffic Analysis Cybersecurity
Beijing University of Posts and Telecommunications Beijing, China · Zhongguancun Laboratory Beijing, China
More Than Meets the Eye: A Semantics-Aware Traffic Augmentation Framework for Generalizable Website Fingerprinting
May 12, 2026Youquan Xian, Xueying Zeng, Lingjia Meng +6Synthetic Data Augmentation Cross-Domain Generalization
School of Cyberspace Security, Beijing University of Posts and Telecommunications, Beijing, China · School of Computer Science and Engineering, Beihang University, Beijing, China · Zhongguancun Laboratory, Beijing, China +2
FlowAtom: Atom-Based Evidence Aggregation for Multi-Label Website Fingerprinting
Sep 24, 2026Chongru Fan, Wentao Huang, Wei Wang +4Multi-Label Classification Prototype Learning
School of Cyberspace Security, Beijing University of Posts and Telecommunications, Beijing, China · Zhongguancun Laboratory, Beijing, China

Jun 16, 2026J/K move · Enter open · S save

ResAware: Cross-Environment Website Fingerprinting via Resource-Privileged Distillation

Chongru Fan, Wei Wang, Wentao Huang, Zhenquan Ding+4

Beijing University of Posts and Telecommunications Beijing, China · Zhongguancun Laboratory Beijing, China

Encrypted Network Traffic Analysis Cybersecurity Learning Using Privileged Information

Open paper AskPDF arXiv

While Website Fingerprinting (WF) attacks achieve high accuracy in controlled laboratory settings, they often degrade substantially in real-world environments due to spatio-temporal drift, browser heterogeneity, proxy obfuscation and etc. This limitation stems from their sole reliance on low-level traffic features that are noisy and highly sensitive to environmental perturbations. To address this problem, we propose ResAware, a cross-environment resource-aware distillation framework under a training-rich/inference-poor asymmetric setting. Specifically, ResAware trains a teacher model on resource-level features, and then distills the resulting privileged knowledge into a student model through heterogeneous knowledge distillation. At deployment time, the student model performs inference using only encrypted traffic, incurring zero additional cost. We evaluate ResAware on a large-scale dataset collected over five months from six globally distributed vantage points, comprising more than

160{,}000

paired samples. The results show that ResAware significantly enhances the cross-environment robustness of diverse WF baselines. Under a 150-day temporal drift, for example, ResAware improves the F1-score of Var-CNN from

72.77\%

81.49\%

and the open-world

TPR@1\%FPR

from

22.40\%

27.20\%

. Our results demonstrate that resource-level supervision improves WF robustness without expanding online observation capabilities.

Abstract

Encrypted Network Traffic Analysis

Explore similar work

ResAware: Cross-Environment Website Fingerprinting via Resource-Privileged Distillation

More Than Meets the Eye: A Semantics-Aware Traffic Augmentation Framework for Generalizable Website Fingerprinting

FlowAtom: Atom-Based Evidence Aggregation for Multi-Label Website Fingerprinting