View on GitHub

explainwf-popets2023

Data-Explainable Website Fingerprinting with Network Simulation

[back to homepage]

Datasets

In this file we describe the datasets that are made available in this artifact.

To access the files, you can get a local copy of the artifact (git clone git@github.com:explainwf-popets2023/explainwf-popets2023.github.io.git).

Fidelity

For the fidelity experiments in Section 3 of the paper, we have data from live Tor crawls and data from the Shadow experiments. In both cases, the urls.txt file contains a list of the URLs from the wikipedia mirror that were fetched during the experiment.

Live Tor

The web pages were crawled using TBB and wget2. The traces resulting from these crawls is located in the fidelity/livetor subdirectory.

Shadow

We ran Shadow simulations fetching the URLs using wget in Shadow. We repeated the simulation six times in order to gather enough traces for analysis. The traces resulting from these crawls is located in the fidelity/tornet-net_0.25-load_2.0 subdirectory. The data from these experiments can be combined into a single file for analysis purposes using the fidelity/tornet-net_0.25-load_2.0/concat_all.sh script.

Sensitivity and Robustness

For the sensitivity and robustness experiments in Sections 4 and 5 in the paper, we have a total of nine unique Shadow networks. We ran two simulations in each network to obtain additional traces for analysis. The traces resulting from these simulations are located in subdirectories of the sensitivity_robustness subdirectory. The data from these experiments can be organized for analysis purposes using the sensitivity_robustness/concat_traces.sh script.