ProbeTags: Privacy-preserving proximity detection using Wi-Fi management frames

Since the beginning of the ubiquitous computing era, context-aware applications have been envisioned and pursued, with location and especially proximity information being one of the primary building blocks. To date, there is still a lack of feasible solutions to perform proximity tests between mobile entities in a privacy-preserving manner, i.e., one that does not disclose one´s location in case the other party is not in proximity. In this paper, we present our novel approach based on location tags built from surrounding Wi-Fi signals originating only from mobile devices. Since the set of mobile devices at a given location changes over time, this approach ensures the user´s privacy when performing proximity tests. To improve the robustness of similarity calculations, we introduce a novel extension of the commonly used cosine similarity measure to allow for weighing its components while preserving the signal strength semantics. Our system is evaluated extensively in various settings, ranging from office scenarios to crowded mass events. The results show that our system allows for robust short-range proximity detection while preserving the participants´ privacy.