2025 In progress Tools for Automated Protocol Analysis Supervisor: Nils Rollshausen What's the state of the art in automatic analysis of unknown binary payloads / protocols? How can we design usable tools to aid researchers in understanding them?
2025 Completed Hiding in Plain Sight: LLM Steganography for Instant Messaging Supervisor: Nils Rollshausen Large Language Models as "next word prediction engines" can be used in interesting ways to embed secret messages in plausible-looking cover-text. This thesis investigates using LLMs for steganography in the context of an Android instant messaging app.
2025 Completed Fuzzing the Unfuzzable? A Comparative Analysis of Fuzzing Methods for Apple Watch Communication Protocols Supervisor: Nils Rollshausen The Apple Watch, like other IoT devices, uses a host of different bespoke wireless protocols. How can we use fuzzing to systematically test Apple's implementation for security issues?
2024 Completed Security Analysis of a Kids Smartwatch Supervisor: Nils Rollshausen Tech companies are increasingly marketing smartwatches to children. But are the manufacturers taking appropriate measures to protect children's health and location data? This thesis investigates the popular Xplora X6 play smartwatch.
2024 Available now Assorted Topics in Wearable Security & Adversarial Interoperability Supervisor: Nils Rollshausen Are you interested in wearable devices, reverse engineering, security research, mobile ecosystems, and interoperability? I might just have a topic for you. Possible topics could be: Explore cross-network collaborative content blocking on social media Create hardware gadgets for adversarial interoperability between "incompatible" smart ecosystems Reverse-engineer & analyze interesting aspects of Apple's ecosystem Explore how to analyze unknown binary data Create an amazing dataset to benchmark reverse-engineering tools Whether you're interested in one of those topics or have an idea of your own in this area, get in touch and we can have a chat.
2024 27th ACM Conference on Computer-Supported Cooperative Work and Social Computing (CSCW 2024) Conference PairSonic: Helping Groups Securely Exchange Contact Information Florentin Putz Steffen Haesler Thomas Völkl Maximilian Gehring Nils Rollshausen Matthias Hollick PDF BibTeX DOI: 10.1145/3678884.3681818 Abstract Securely exchanging contact information is essential for establishing trustworthy communication channels that facilitate effective online collaboration. However, current methods are neither user-friendly nor scalable for large groups of users. In response, we introduce PairSonic, a novel group pairing protocol that extends trust from physical encounters to online communication. PairSonic simplifies the pairing process by automating the tedious verification tasks of previous methods through an acoustic out-of-band channel using smartphones' built-in hardware. Our protocol not only facilitates connecting users for computer-supported collaboration, but also provides a more user-friendly and scalable solution to the authentication ceremonies currently used in end-to-end encrypted messengers like Signal or WhatsApp. PairSonic is available as open-source software: https://github.com/seemoo-lab/pairsonic
