README.md

Algorithm Collection

This repository aggregates multiple network traffic classification projects. Each subproject has its own README with detailed usage. This top-level README provides a quick overview and entry points.

Repository Structure

• Fs-net/

  • BiGRU-based sequence model for encrypted traffic classification using packet-length sequences.
  • See Fs-net/README.md for details.

• 1d-CNN/

  • PyTorch implementation of a 1D-CNN-style convolutional classifier trained on gzipped ubyte-format datasets (12/6/2-class).
  • See 1d-CNN/README.md for details.

• TFE-GNN/

  • Provides full preprocessing from pcap to byte-level traffic graphs, training and evaluation scripts.
  • See TFE-GNN/README.md for details.

• AppSniffer/

  • VPN traffic detection and VPN app classification using AutoGluon; supports single-dataset training/testing and cross-validation.
  • See AppSniffer/README.md for details.

• Rosetta/

  • Enable robust TLS encrypted traffic classification in diverse network environments with TCP-aware traffic augmentation.
  • See Rosetta/README.md for details.

Notes

References

• 1d-CNN paper : https://dl.acm.org/doi/10.1109/ISI.2017.8004872

• Fs-net paper : https://ieeexplore.ieee.org/document/8737507

• TFE-GNN paper: https://dl.acm.org/doi/abs/10.1145/3543507.3583227

• AppSniffer paper: https://dl.acm.org/doi/10.1145/3543507.3583473

• Rosetta paper: https://dl.acm.org/doi/10.1145/3603165.3607437