Research On Privacy Protection Methods For Resisting Trajectory Similarity Attacks

With the arrival of the era of big data,people began to mine and use massive data,location-based services have been widely used and achieved good results.However,LBS provides services on the premise of accessing users' real-time location,and the service process may violate users' location or other personal privacy.Especially in the case of continuous LBS queries,there may be potential malicious attackers in service providers,which can use background knowledge to infer user location and trajectory privacy.At present,a lot of privacy leaks have been reported,people pay more and more attention to their privacy.How to protect users' privacy effectively is becoming more and more important.This paper aims at the shortcomings of the existing location-based privacy protection and trajectory privacy protection schemes.The main work is as follows:With the arrival of the era of big data people begin to mine and use massive data.Location-based services(LBS)are applied more and more widely.However,the premise of location-based services(LBSP)is to access users' real-time location,and the service process may infringe on users' location or other personal privacy.Especially in continuous LBS queries,there may be malicious attackers in service providers who use background knowledge to infer user location and trajectory privacy.At present,a large number of privacy leakage incidents have been reported,and people continue to pay more attention to their privacy.How to effectively protect users' privacy is becoming increasingly important.This paper aims at the shortcomings of the existing location-based privacy protection and trajectory privacy protection schemes.The main work is as follows:(1)Analyzing the existing LBS service location privacy protection methods,a collaboration-based personalization(k,r,s)-privacy protection scheme is proposed.This scheme is mainly aimed at continuous query services.By constantly searching for collaborative users,sharing available caching information,and using cached data to satisfy query requests.It reduces the information exchange between the query user and LBSP in continuous location query,makes LBSP unable to obtain enough location information to reconstruct the query user's trajectory,and effectively prevents the differential recognition attack.(2)Researching the existing anonymity model of trajectory data publishing,the trajectory anonymity model resisting trajectory similarity attack is proposed.Trajectory(k,e)-anonymity model mainly improves the traditional trajectory k-anonymity,and solves the problem of trajectory privacy leakage caused by high trajectory similarity in anonymous set.The innovation of trajectory(k,e)-anonymity model is that the trajectory slope is regarded as the trajectory sensitive value,and at least k trajectories with different slopes are grouped into one group,and the difference of trajectory slope in each class is required to be at least e.The purpose of this paper is to reduce the possibility of trajectory privacy disclosure when users use LBS services,and to reduce the possibility of trajectory privacy disclosure when trajectory data is published.In LBS continuous query,this algorithm can effectively reduce the amount of real information sent to LBSP,prevent trajectory reconstruction,and reduce the probability of location and trajectory privacy leakage by cooperating users to share cached data to meet user query service requests.In trajectory data publishing,the traditional trajectory k-anonymity model is improved to reduce the trajectory similarity within the trajectory equivalent class,effectively resist the trajectory similarity attack,and the degree of privacy protection is significantly improved.