A Trajectory Privacy Protection Scheme Based on the Replacement of Stay Points

Location-based services generate a large amount of location and trajectory data, which contain rich spatiotemporal and semantic information. Publishing these data without proper protection can seriously threaten users’ trajectory privacy. Existing trajectory privacy protection schemes generally fail to consider the dependency between a stay point and its preceding location and also overlook the relationship between the semantic information of location and privacy. Moreover, they often suffer from issues such as over-protection. Therefore, this paper proposes a trajectory privacy protection scheme based on the replacement of stay points. First, a stay point extraction algorithm is proposed, which extracts users’ stay points by setting distance and time thresholds based on the principle of the sliding window. Then, this paper proposes a location perturbation algorithm based on the vector indistinguishability mechanism and introduces different protection strategies for ordinary stay points and long-duration stay points, respectively. Finally, the perturbed trajectory is adjusted by generating a certain number of location points near the replacement points to maintain the temporal continuity and integrity of the trajectory. The experimental results indicate that it is necessary to provide more meticulous protection for long-duration stay points. Compared with similar schemes, the proposed scheme in this paper achieves higher data utility while ensuring privacy.