Mobile Systems Location Privacy: "MobiPriv" A Robust K Anonymous System

With the rapid advancement of positioning and tracking capabilities (mobile phones, on-board navigation systems) location based services are rapidly increasing. Privacy in location based systems is addressed in many papers. Our work is focused on the trusted third party privacy framework that utilizes the concept of k-anonymity with or without I-diversity. In previous anonymization models k may be defined as a personalization parameter of the mobile user or as uniform system parameter for all mobile users. Clearly, k other users may not be available at the time of request in these systems. These requests are discarded because the quality of service (QoS) they require cannot be satisfied. In this paper we introduce a novel suite of algorithms called MobiPriv that guarantees a 100% success rate of processing a mobile request using k-anonymity with diversity considerations. We evaluated our suite of algorithms experimentally against previously proposed anonymization algorithms using real world traffic volume data, real world road network and mobile users generated realistically by a mobile object generator.