Edge computing-based big data privacy preservation in motion trajectory prediction for martial arts training

Abstract

Abstract The development of big data technology has brought about sweeping changes in many fields. Privacy protection has become a key concern for big data users. The study constructs a geographically indistinguishable location-based privacy protection mechanism based on differential privacy and geographic indistinguishability and further constructs an edge computing-based privacy protection model for martial arts movement trajectories. The performance of an edge computing-based privacy protection model for martial arts sports trajectory is examined by comparing it to other models in terms of quality of service loss, privacy protection strength, and range counting queries. And explore the effects of service type and number of users on the system. The perturbation distance on all three time periods of the Geo-In method proposed in this paper is the smallest among the five methods, and the perturbation distance decreases with the increase of the privacy budget, and the loss of quality of service is minimized. The Geo-In algorithm’s perturbed locations have less semantic similarity to the real locations of martial arts training, resulting in stronger privacy protection of the locations. Range counting queries experience a decrease in relative error as the number of users and query range increase.