About Me
In lieu of a photo of me, please enjoy this crow.
Image by Alexas_Fotos
Hi, I'm Vecna, a PhD student doing privacy-related computer science research at the University of Waterloo. I am part of the Cryptography, Security, and Privacy research group, and my supervisor is Ian Goldberg.
I am passionate about privacy, free software, and other topics at the intersection of ethics and computer science. Computer science is inherently political, and it is the obligation of computer scientists to work on making the world better and not worse. In my research area, this means empowering people with privacy-enhancing technologies rather than building systems that enable mass surveillance by governments and corporations.
What to Call Me
- Name: Vecna
- Pronouns: they/them
- Gendered terms: no
You Might Know Me As...
- The one who sends the CrySP speaker schedule emails (Assistant Vice Speaker Rotator)
- The one who runs the CrySP Matrix server (Transadmin)
- Your classmate
- A TA for your class (probably CS 459)
Research
I work on privacy and anti-censorship research.
Current and Upcoming Projects
Troll Patrol: Detecting Blocked Tor Bridges
Troll Patrol is a system for detecting which Tor bridges (especially those distributed by Lox) have been blocked by censors. This was the topic of my magisterial thesis (see below), which produced the basic system and evaluated it in simulation. TODO: Further practical evaluation and refinement as necessary.
Key verification user study
I am interested in exploring key verification in secure messaging apps (such as Signal). In particular, can we incentivize verification of users' safety numbers, and (if so) does this meaningfully improve security? (Do users understand the implications?)
Preprints
Random (Un)rounding : Vulnerabilities in Discrete Attribute Disclosure in the 2021 Canadian Census
Chris West, Vecna, Raiyan Chowdhury. arXiv preprint arXiv:2307.13859.
[PDF]
Brief summary: The primary privacy technique used in the 2021 Canadian Census, "random rounding", probabilistically rounds counts to multiples of 5. We demonstrate cases in which original counts can be recovered with certainty and cases in which original counts can be inferred (albeit with some probability of guessing incorrectly).
Theses
[MMath] Troll Patrol: Detecting Blocked Tor Bridges
Magisteriate in Mathematics Thesis, University of Waterloo, 2024.
[PDF]
Brief summary: Secret proxies such as some types of bridges may be used to circumvent censorship in countries where the use of Tor is forbidden. Lox, a reputation-based bridge distribution system, requires knowledge of which bridges have been blocked by censors and which have not. We present a system for detecting censorship of Tor bridges based on usage statistics from those bridges and anonymous reports that users can submit when they are unable to connect to their bridges.
Impact: The Troll Patrol code from this work has been merged into the experimental Lox codebase (but not actually deployed).
