This repository contains the source code and data for Stream Fingerprinting using nerual networks. The attack is an encrypted traffic analysis attack that allows a passive adversary to infer which YouTube video a user watches by sniffing encrypted traffic between a user and the server.
The dataset and code are for research purposes only. The results of this study are published in the following paper:
Haipeng Li, Ben Niu, Boyang Wang, “SmartSwitch: Efficient Traffic Obfuscation against Stream Fingerprinting,” 16th EAI International Conference on Security and Privacy in Communication Networks (SecureComm 2020), October, 2020.
The video_collection directory contains the code to automatically collect stream traffic between user and stream provider (e.g. Youtube).
The feature_selection directory contains the code for feature selection. We have two different catogeries of selection methods, Mutual Information based methods and SW-PFI, which is proposed in our paper.
The defense directory contains the defense code for generating obfuscated stream traffic.
We investigated 100 classes (i.e., Youtube videos) and 200 traffic traces per class in this research. The original traffic data (i.e., non-defended traffic data), defended data and the list of YouTube videos we used in this study can be found below (last modified: May, 2022):
Note: the above link needs to be updated every 6 months due to certain settings of OneDrive. If you find the link is expired and you cannot access the data, please feel free to email us ([email protected]). We will be update the link as soon as we can. Thanks!
We leveraged a Convolutional Neural Network to infer which video it is based on the traffic pattern. The CNN includes 11 layers and achieved over 90% accuracy in the attack. Details of the structure and tuned hyperparameters can be found in our paper.
When reporting results that use the dataset or code in this repository, please cite:
Haipeng Li, Ben Niu, Boyang Wang, “SmartSwitch: Efficient Traffic Obfuscation against Stream Fingerprinting,” 16th EAI International Conference on Security and Privacy in Communication Networks (SecureComm 2020), October, 2020.
Haipeng Li, [email protected], University of Cincinnati
Boyang Wang, [email protected], University of Cincinnati