SPIRAL: A Scalable Private Information Retrieval Approach to Location Privacy

Abstract

Protecting users' location information in location-based services, also termed location privacy, has recently garnered significant attention due to its importance in satisfying users' privacy concerns when using location-aware services. Several approaches proposed in the literature blur the user's location in a region by increasing its spatial extent or anonymizing the user among several other users. Such approaches in nature require users to communicate through a trusted anonymizer for all of their queries which can impose unrealistic overall communication/computation overhead between the server and the anonymizer for users with more stringent privacy requirements. We revisit the location privacy problem with the objective of providing significantly more stringent privacy guarantees and propose SPIRAL, a scalable private information retrieval approach to location privacy, which is to the best of our knowledge, the first approach to utilize practical private information retrieval (PIR) as a more fundamental approach to enable blind evaluation of range queries. We perform several experiments on real-world data to evaluate the effectiveness and the feasibility of our approach.