Trajectory Privacy-Preserving Approach for Consecutive Queries Based on K-Anonymity

Existing approaches to protecting the privacy of trajectories using k-anonymity when requesting LBSs. However, the problem of anonymization is even harder due to dependencies between consecutive queries. In order to solve problem, a novel trajectory privacy-preserving approach based on k-anonymity is used in this paper. By formulating a query region, a spatial index model of trajectory data is established that can be further used by continuous KNN (K Nearest Neighbor) method to compute a candidate set of trajectory that have the similar area with current trajectory. In order to find the optimal anonymity path from k trajectories, a heuristic based approach is proposed. Empirical results show that the proposed method significantly decreases information losses for the same privacy levels and that information distortion strongly depends on value of parameter k. The k-anonymity based approach can effectively solve the problem arose from dependencies between consecutive queries and strengthen the privacy of trajectory data.