DANE is a hackable dataset generation tool to collect network traffic in a variety of configurable network conditions.
It runs on Windows, Mac, and Linux.
Table of contents
DANE provides two core functionalities:
-
Automatically collect network traffic datasets in a parallelized manner
Manual data collection for network traffic datasets is a long and tedious process—run the tool and you can easily collect multiple hours of data in one hour of time (magic!) with one or many desired 'user' behaviors.
-
Emulate a diverse range of network conditions that are representative of the real world
Data representation is an increasingly relevant issue in all fields of data science, but generating a dataset while connected to a fixed network doesn't capture diversity in network conditions—in a single file, you can configure DANE to emulate a variety of network conditions, including latency and bandwidth.
You can easily hack the tool to run custom scripts, custom data collection tools, and other custom software dependencies which support your particular research interest.
For all documentation, including a quick start, details about the technical approach, and FAQs, please consult the website 📖.
https://dane-tool.github.io/dane
See something you'd like improved? Better yet, have some improvements coded up locally you'd like to contribute?
We welcome you to submit an Issue or make a Pull Request detailing your ideas!
This project was originally created in affiliation with the Halıcıoğlu Data Science Institute's data science program at UC San Diego.
https://hdsi.ucsd.edu/, https://dsc-capstone.github.io/
DANE was motivated and developed with the generous support of Viasat.
https://viasat.com/