Unfettered access to the Internet is unfortunately not universal—recent studies show that more than half of the world’s population is subject to at least some censorship. Even in regions without censorship, Internet outages (e.g., during natural disasters) can hinder the ability to communicate online. Avoiding censorship and communicating during Internet outages have inspired a number of proposals for communicating via a class of decentralized routing protocols based on {}em sneakernets}. In a sneakernet, messages are passed between human-carried devices (usually, smartphones), entirely avoiding the Internet. Importantly, the movement of messages in a sneakernet is dictated by the movements of the (human) device owners; these networks tend to be opportunistic in the sense that messages are exchanged between parties only when those parties encounter one another through happenstance. Understanding the security, performance, and privacy properties of various sneakernet protocols remains an open problem, with individual proposals inventing their own metrics and evaluation methodology. This paper presents Cadence, a simulator for decentralized human movement-based communication protocols that provides a unifying environment for evaluating sneakernet protocols. We describe the architecture of Cadence, argue its benefits to network and security researchers, and present use-cases for studying proposed sneakernet protocols. Cadence is free open-source software that will be released with the publication of this paper.
@inproceedings{berger_cadence_2023,
title = {Cadence: {A} {Simulator} for {Human} {Movement}-based {Communication} {Protocols}},
booktitle = {Workshop on {Cyber} {Security} {Experimentation} and {Test} ({CSET})},
author = {Berger, Harel and Sherr, Micah and Aviv, Adam J.},
month = aug,
year = {2023}
}