Research On Privacy Preserving Trajectory Data Publishing

In recent years, with the rapid development of wireless communication technology, user can use the mobile sevices whenever and wherever they want (such as mobile phones, tablet computers, etc.)to access to information and services. And with the widespread use of GPS,the rapid development of location positioning devices (such as mobile phones, GPS navigation, etc.),A class of Location Based Service based on user’s location information came into being.However, with enjoying the location-based services bring the convenience,users also faces the risk of disclosure of personal privacy. In order to obtain location-based services, user needs to send the personal location information to the server provider. If the server provider is not secure, the user’s privacy is facing the risk of being leaked. Because the server provider can collect the user’s location information to create a user’s trajectory data, and then publish the trajectory data to a third party institution to do some trajectory data mining and analysis in order to support a variety of applications which related. Therefore, the user’s location privacy protection is not enough. How to protect user’s privacy in the maximum possible under trajectory data publishing has become a hot issue of research scholars.This paper focuses on a number of privacy issues under the presence of trajectory data publishing. Specific content includes the following three points:(l)Under the k-anonymity model, there may exist some trajectories with a high degree of similarity being covered in a same anonymous set, and resulting the anonymous set may still retain the sensitive information which in the original trajectories.To solve this problem, the paper through dividing the two-dimensional space into grids of equal size.And defined the 1-diversity in trajectory data publishing. Under the k-anonymity model and l-diversity,this paper design a greedy algorithm based on clustering to structure anonymity set. Based on real data sets Simulation results show the feasibility and effectiveness of the proposed algorithm.(2)Depth study on the κ-anonymity model under the trajectory privacy, Current trajectory privacy protection algorithm based on κ-anonymity model assumes that all users have the same degree of privacy protection k. In real life, privacy protection needs of different individuals are often different, use the largest degree of privacy protection k to structure κ-anonymity privacy protection set will result in reduced availability of anonymous data. In response to these deficiencies, this paper proposes trajectory privacy protection algorithms support the personalized trajectory Privacy protection requirements. Based on real data sets Simulation results show the feasibility and effectiveness of the proposed algorithm.(3)Under the κ-Anonymity model, current most of the trajectory data privacy protection algorithms are using clustering and other methods divide the trajectory data to structure anonymity set.However, when a trajectory data is long and the spatial radiation field which it across is large, often leads to a larger anonymous space that affect the data availability. In response to these deficiencies,this paper propose a trajectory segmentation anonymity algorithm based the grid. And for the privacy risk under the overlap attack.This paper design a trajectory anonymity algorithm base on the space divided under the quadtree. Based on real data sets Simulation results show the feasibility and effectiveness of the proposed algorithm.