Mixed Reality (MR) platforms are becoming increasingly integrated into daily life, raising concerns about privacy and security. Centralized systems such as Meta’s Horizon
often expose sensitive user data to a central authority, creating risks of surveillance and loss of control. Alternative approaches, such as distributed or middleware-based MR
infrastructures, reduce these risks but still face challenges in enforcing privacy policies and enabling secure contextual interactions.
This thesis investigates privacy-preserving functions in MR with a focus on publish/subscribe communication. We review cryptographic methods such as Secure Multi-Party Computation (SMPC), Private Set Intersection (PSI), and Homomorphic Encryption (HE), analyzing their potential for contextual functions like proximity detection. The trade-offs between privacy, latency, and feasibility are discussed.
In the practical part, we extend the Ubii-Interact framework with a privacy-preserving publish/subscribe layer. Using SMPC, we implement a prototype for distance-threshold proximity detection, ensuring that sensitive spatial coordinates remain private while still
enabling contextual sharing. Evaluation shows that the prototype achieves correct results with interactive latency, demonstrating the feasibility of integrating cryptographic privacy mechanisms into MR middleware.
Overall, this work contributes to designing privacy-aware MR environments and offers guidance for integrating secure contextual services into real-world systems.