A Random Anonymity Framework for Location Privacy

摘要

Location k-anonymity techniques typically use anonymous spatial region to ensure privacy. But these solutions are vulnerable to multiple queries attacks and inference attacks. Failing to account for the obstacle in geographic space is a severe problem since adversaries will surely regard these constraints. A novel framework is proposed to enhance location-dependent queries, based on the theoretical work of k-anonymity and Voronoi diagrams, allows a user to express service requirement and privacy requirement by specifying a region and an appropriate value of k. A trusted anonymity server form a restricted set(k, r, s), which is composed of a number of discrete points to meet the requirements for location k-anonymity and location l-diversity. The location-based services(LBS) server implements an efficient algorithm for continuous-region-query processing. Simulation results demonstrated that the framework is superior to previous works in terms of privacy. Moreover, discreteness and randomness of the anonymous set are conducive to resisting location tracking attacks.