CipherSpectrum
A dataset of TLS 1.3 encrypted network traffic designed for network traffic analysis, introduced in the paper: “SoK: Decoding the Enigma of Encrypted Network Traffic Classifiers” [IEEE Symposium on Security and Privacy (S&P) - 2025].
📥 Download
The CipherSpectrum dataset has 4 ZIP files, each containing network traffic data for the same set of 40 domains as follows:
- 📦 aes-128-gcm.zip — Network traffic encrypted with TLS_AES_128_GCM_SHA256
- 📦 aes-256-gcm.zip — Network traffic encrypted with TLS_AES_256_GCM_SHA384
- 📦 chacha20-poly1305.zip — Network traffic encrypted with TLS_CHACHA20_POLY1305_SHA256
- 📦 mix.zip — A balanced mixture of sessions encrypted using all three TLS 1.3 mandated cipher suites
📌 Overview
Each traffic trace was collected by visiting the selected domains using customized Chromium, and Firefox browsers.
| Attribute | Value |
|---|---|
| Total Classes | 40 |
| Cipher Suites | 3 (TLS-AES-128-GCM, TLS-AES-256-GCM, TLS-CHACHA20-POLY1305) |
| Sessions per Class per Suite | 1,000 |
| Total Sessions | 120,000 |
| Collection Period | Jan – Mar 2024 |
| Customized Chromium Source Files | Github Repository |
🔐 Why CipherSpectrum?
Most publicly available NTC datasets:
- Contain unencrypted or outdated traffic
- Include traffic sessions encrypted with deprecated cipher suites (e.g., RC4, 3DES)
- Suffer from class imbalances or limited raw traffic
CipherSpectrum addresses these issues by providing equal representation for each class, using TLS 1.3 traffic encrypted with AEAD cipher suites. For more information, please refer to our paper.
🧪 Research paper
For more details on the CipherSpectrum dataset, refer to our research paper:
IEEE Xplore IEEE CSDL Preprint📖 Citation
If you use CipherSpectrum in your research, please cite:
@inproceedings{,
author = { Wickramasinghe, Nimesha and Shaghaghi, Arash and Tsudik, Gene and Jha, Sanjay },
booktitle = { 2025 IEEE Symposium on Security and Privacy (SP) },
title = {{ SoK: Decoding the Enigma of Encrypted Network Traffic Classifiers }},
year = {2025},
ISSN = {2375-1207},
pages = {1825-1843},
doi = {10.1109/SP61157.2025.00165},
url = {https://doi.ieeecomputersociety.org/10.1109/SP61157.2025.00165},
publisher = {IEEE Computer Society},
address = {Los Alamitos, CA, USA},
month = {May}
}
License & Terms
The CipherSpectrum dataset is licensed under the CC BY-NC 4.0 License.
Free use of the CipherSpectrum dataset is hereby granted in perpetuity for academic and non-commercial research purposes. Use of this dataset for commercial purposes requires prior permission from the authors.
📧 Contact
For any questions or feedback, please contact
🔍 Interested in what we do?
Find out more about how we bridge the gap between network intelligence and AI:
NTC-IQ