Masking Location Streams in the Presence of Colluding Service Providers

Abstract

Many service providers rely on location data or streams to offer personalized services. However, reckless release of location streams can have serious privacy implications, especially with respect to sensitive zones, paths and time intervals. Unfortunately, current approaches by mobile platform providers to enhance privacy expose major shortcomings, especially with respect to releasing location information of frequently visited areas and in the presence of colluding service providers that are willing to combine local knowledge to retrieve more accurate information.This paper presents a practical approach to obfuscate location streams in the presence of the aforementioned challenges. Our solution combines multiple technologies and is validated through the development of a novel privacyfriendly location service in Android. It aims to realize a reasonable trade-off between privacy and utility concerns respectively raised by end-users and service providers. Thereby, sensitive locations, zones, paths, timestamps, and time intervals are highly configurable.