This repository contains a Loopix mix network simulator that accompanies the paper accepted at USENIX Security 2021: "Rollercoaster: An Efficient Group-Multicast Scheme for Mix Networks".

The simulator and tooling allow:

• Creating new simulation configurations via a Jupyter Notebook.
• Running these configurations via a deterministic simulator.
• (Re-)Creating the graphs presented in the paper.

The software in this repository is licensed under the MIT license as described in the LICENSE file.

Optionally: For large simulation configurations it contains helper scripts for scheduling the simulator jobs on a computing cluster that can execute slurm files.

Making effective use of this repository requires at least the following:

• Basic usage of standard developer tools: command line, git, text editor
• Basic understanding of Python
• Starting and editing Jupyter Notebooks
• Creating and running Docker images

The artefact runs on UNIX like systems with a recent version of Docker installed. The simulations and the resulting data can be large - especially when loaded in Jupyter Notebooks. I suggest using a machine with at least 16 GiB RAM (preferably more) and 10 GiB of free disk space. The code has been tested on Ubuntu 20.04 and Mac OS 11.3.

NOTE FOR AEC: If you do not have access to a Unix-like system with 16GiB, I can provide you with remote access to a server with such capabilities.

Please read the files within the docs/ folder before starting experimenting with the code. The walkthrough guide is the main document for the artefact evaluation.

• architecture.md file explains the high-level view of this artefact. This provides understanding of its structure and provides orientation if you are getting lost.
• walkthrough.md is a tutorial with a detailed step-by-step description for reproducing the figures from the paper.