A context-dependent privacy preserving framework in road networks

Abstract

The prevalence of Location Based Services (LBSs) increases personal privacy concerns due to the untrustworthy service providers. We demonstrate a context-dependent privacy preserving framework for users whose movements are confined by the underlying road networks. Both the location privacy and query privacy in continuous queries are preserved as they are closely related. For continuous query services, different positions on a user's trajectory may have different privacy sensitivities. In addition, privacy is about users' feelings and varies among them. Hence, a Policy Service(PS) is introduced to generate context-dependent privacy strategies according to user-defined privacy profiles. Meanwhile, a semi-honest Anonymizing Service(AS) is employed to generate prediction-based cloaks with history information for users while satisfying their privacy strategies. The PS and AS interact with each other in the way to ensure neither of them can obtain both the location information and the query contents. The simulated results show the effectiveness of our framework in the view of privacy preserving and system performance.