CrySP PhD Seminar: Website Fingerprinting Talk
- University of Waterloo’s Cryptography, Security, and Privacy group scheduled a PhD seminar for Tuesday, April 28, 2026, where candidate Michael Wrana will present TSA-WF, his work on website fingerprinting. - Wrana’s abstract says TSA-WF keeps packet timing and direction intact so researchers can test classical time-series similarity methods on encrypted traffic, instead of relying only on newer machine-learning approaches. - The talk follows a 2025 ARES paper by Wrana, Uzma Maroof, and Diogo Barradas on the same method, placing the seminar inside active privacy-attack research. (crysp.uwaterloo.ca)
University of Waterloo’s CrySP group will host an online PhD seminar on Tuesday, April 28, 2026, where Michael Wrana presents research on website fingerprinting. (cs.uwaterloo.ca) Website fingerprinting is a traffic-analysis attack: an observer studies the shape of encrypted network traffic to infer which site a person visited, even without reading the page contents. (cs.uwaterloo.ca) Wrana’s talk is titled “Exploring the Effectiveness of Time Series Analysis for Website Fingerprinting,” and the event listing says it runs from 9:00 a.m. to 10:00 a.m. Eastern Daylight Time. (cs.uwaterloo.ca ) (cs.uwaterloo.ca) The core idea is to treat traffic traces like a time series — a sequence ordered over time — and compare their patterns the way analysts compare other signals, such as sensor readings or stock movements. (cs.uwaterloo.ca) The abstract says Wrana introduces a pipeline called TSA-WF that is designed to preserve packet timing and packet direction, two details that can define a browsing session’s observable shape. (cs.uwaterloo.ca) That setup lets researchers test “algorithms designed to measure time series similarity” in the website-fingerprinting setting, according to the seminar page. (cs.uwaterloo.ca) The seminar page identifies Wrana as a PhD candidate at the David R. Cheriton School of Computer Science and lists Diogo Barradas as his supervisor. (cs.uwaterloo.ca) CrySP’s publications page shows the project already appeared as a 2025 paper: “TSA-WF: Exploring the Effectiveness of Time Series Analysis for Website Fingerprinting,” by Michael Wrana, Uzma Maroof, and Diogo Barradas, at ARES 2025. (crysp.uwaterloo.ca) The event is listed as online in both the seminar page and the school’s events listings, which place it among the Cheriton School’s regular PhD seminars. (cs.uwaterloo.ca 1) (cs.uwaterloo.ca 2) For CrySP, the April 28 seminar is a public snapshot of a live research question: how much encrypted traffic still reveals when timing patterns are preserved and compared carefully. (cs.uwaterloo.ca)
