Research On Privacy Protection Technology Of Trajectory Data Of Moving Objects

With the rapid development of GPS technology and mobile devices,location-based services,as a general technical application,have played an important role in personal social interaction,social progress and national development.Data publishing provides the most basic support for data mining.When people enjoy the convenience brought by this location-based service and data mining technology,a large amount of positioning data has been collected and mined,and has even fallen into the hands of illegal attackers.As a special form of positioning data,trajectory data contains a large amount of private information of moving objects.When massive trajectory data is released directly without processing,it can bring more convenience,but at the same time,the risk of personal privacy leakage is huge.Therefore,it is necessary to conduct privacy protection research on trajectory data of moving objects.The existing methods for privacy protection of trajectory data usually generalize,anonymize and add noise on the original trajectory data.This thesis is based on the k-anonymity technology of existing research,aiming at the excessive anonymity and neglect of current research methods.The characteristics of the behavior patterns of moving objects cause the problems of low availability and poor privacy protection effect after data release.Two privacy protection methods for moving object trajectory data are proposed,which are different from existing research.(1)Aiming at the problem of low availability of anonymously processed trajectory data in sparse environment,this thesis proposes a trajectory privacy protection method based on fake trajectories in sparse environment.In this algorithm,considering the geographical environment where the trajectory data of the moving object is located,the k-anonymity technology is combined with the suppression method,and reasonable suppression rules are designed,and the overall direction of the trajectory and the distance of the trajectory are used as an important basis for selecting false trajectories.In addition,it is also proposed to use the access probability defined based on the trajectory dataset to balance anonymity and data availability.In the process of generating false trajectories,a directed graph storage structure is used to fit false trajectories,so as to realize the anonymity of trajectory data.Experiments and analysis are carried out based on the real trajectory data set of moving objects.The experimental results show that the algorithm in this thesis has higher data availability under the condition of satisfying the requirement of trajectory data anonymity.(2)Aiming at the problem that the difference in the behavior patterns of moving objects is too similar to the anonymized trajectory data,a trajectory anonymization method based on equivalent trajectory is proposed.Most of the existing trajectory data privacy protection methods are based on the entire trajectory of the moving object,ignoring the characteristics of the moving object's behavior pattern,movement trend and activity law,and the trajectory with many locations and a large time span is ignored,too large anonymity area directly affects the availability of data after release,and highly similar anonymity sets can easily lead to privacy leakage in extreme cases.The trajectory anonymization method based on equivalent trajectory is mainly aimed at long trajectory.In this method,the original trajectory data set is preprocessed to obtain the equivalent trajectory data set,and then the idea of clustering is used on the equivalent trajectory data set.Constructing anonymity sets with trajectory diversification as a key constraint for constructing anonymity sets,but maintaining similarity in motion trends to improve data availability.The experimental results show that the trajectory data anonymity set constructed by this method has better privacy protection and will not cause a large loss of data information.