Abstract

Location-based services (LBSs) require location information from their clients, yet a person's whereabouts may imply sensitive private information such as heath condition, lifestyle, and political association. In particular, unlike other personal data, location information has the potential to allow an adversary to physically locate a person, and therefore, presents unique and heightened security threats to individuals. For privacy and security protection, a viable solution is to allow users to preserve anonymity in service uses. For this purpose, however, simply using pseudonyms in requesting LBSs is not sufficient, because a user's location information itself may reveal her real-world identity. In this talk, we examine the challenges of supporting anonymous uses of LBSs, and present a novel solution for anonymity protection. At the core of the solution is a suite of location depersonalization techniques that prevent one from identifying a subject based on her location information, either a single location sample or a time-series sequence of them. We discuss the design and implementation of these techniques and the performance results in terms of their impact on the quality of LBSs.