A Personalized Two-Tier Cloaking Scheme for Privacy-Aware Location-Based Services

The ubiquity of modern mobile devices with GPS modules and Internet connectivity such as 3G/4G techniques have resulted in rapid development of Location-Based Services (LBSs). However, users enjoy the convenience provided by the untrusted LBS server at the cost of their privacy. To protect user's sensitive information against adversaries with side information, we design a personalized spatial cloaking scheme, termed TTcloak, which provides k-anonymity for user's location privacy, l-diversity for query privacy and desired size of cloaking region for mobile users in LBSs, simultaneously. TTcloak uses Dummy Query Determining (DQD) algorithm and Dummy Location Determining (DLD) algorithm to find out a set of realistic cells as candidates, and employs a CR-refinement Module (CRM) to guarantee that dummy users are assigned into the cloaking region with desired size. Finally, thorough security analysis and empirical evaluation results verify our proposed TTcloak.