Pseudonym-based anonymity zone generation for mobile service with strong adversary model

The popularity of location-aware mobile devices and the advances of wireless networking have seriously pushed location-based services into the IT market. However, moving users need to report their coordinates to an application service provider to utilize interested services that may compromise user privacy. In this paper, we propose an online personalized scheme for generating anonymity zones to protect users with mobile devices while on the move. We also introduce a strong adversary model, which can conduct inference attacks in the system. Our design combines a geometric transformation algorithm with a dynamic pseudonyms-changing mechanism and user-controlled personalized dummy generation to achieve strong trajectory privacy preservation. Our proposal does not involve any trusted third-party and will not affect the existing LBS system architecture. Simulations are performed to show the effectiveness and efficiency of our approach.