Trajectory Data Protection based on Differential Privacy k-means

Considering that the existing trajectory data protection methods will result in low data availability, this paper considers the characteristics of trajectory data. It proposes a trajectory data protection method that satisfies differential privacy. It not only protects the privacy of the user’s trajectory but also has certain data analyzability. First, with the introduction of the weighted-multi-point judgment method, the infection points in the trajectory are found by setting the threshold of the infection angle. Second, the density of each trajectory point is calculated so as to determine the initial clustering center point. In addition, the privacy protection of the trajectory data is processed with the use of an improved differential privacy k − means method, which improves the availability of the trajectory data on the premise of satisfying the differential privacy. Finally, some numerical experiments are carried out to verify the effectiveness of the method.