User Privacy Protection Method Based on Dynamic Hiding

Abstract

The most commonly used method of location privacy protection is location K-anonymization. At present, most of the K-anonymization models are aimed at the attackers who do not understand the users' background knowledge. The probability of hacking will increase when they know about users. This paper proposes a multi-level meshing method to predict the user's trajectory according to the user's historical track data recorded by the LBS (location based service) server. Then the LBS server determines whether to dynamically adjust the location of the corresponding user in the K-anonymization model K-degree anonymous and remove the redundancy of the anonymous area while satisfying the K-degree anonymous. Due to the increase of the anonymous area, the impact on the quality of the LBS server is reduced. The experiments verify that this method is effective to the privacy protection of users when attackers know about the background knowledge.