/pytorch-pcap-cnn

National Chengchi University 2021 Fall CS Graduate Research Poster, I used this repo. for pcap recog..

Primary LanguagePython

Deep Learning-Based Encrypted Traffic Classification: Novel Use of the Headers

  • This is a Pytorch implementation of 1D-CNN Deep Packet.

  • See 107research_poster.pdf for details.

Dataset

  • This research uses the UNB ISCX VPN-nonVPN dataset.

  • Make sure to download and unzip the dataset. Then run the following command for data preprocess:

    python3 data_preprocess.py {path/to/unzipped/dataset}

  • The preprocessed dataset is saved to ./pre.

Run

  • Make sure you have run the previous steps so that the preprocessed dataset is in ./pre folder.

  • To train the model and see the results, simply run:

    python3 main.py