Studying the anonymity trilemma with a discrete-event mix network simulator

TL;DR

This paper introduces a novel discrete event mix network simulator to analyze how various design choices impact anonymity, latency, and bandwidth overhead in mix networks, supported by empirical analysis of existing systems.

Contribution

First-ever open-source simulator for mix networks enabling detailed analysis of deployment scenarios and design trade-offs.

Findings

Design choices significantly affect anonymity and latency.

Different mix networks exhibit varied scalability and overhead.

The simulator provides insights into optimal configurations.

Abstract

In this work, we present a discrete event mix network simulator, which allows analysing how anonymity, latency, and bandwidth overhead are affected by various scenarios of deployment and design choices. These design choices include network topology, mixing technique, volume of traffic, latency requirements, packet size or use of cover traffic. To the best of our knowledge, this is the first such simulator as work on it began in 2017 to analyze the Loopix mix network, and the code of our simulator is available under an open-source license. To demonstrate the capabilities of our simulator, we perform an empirical analysis of the impact of core design choices on anonymity, scalability and latency in Elixxir, HOPR and Nym, currently deployed mix network infrastructures that make a variety of different choices in their design.