Research On Trajectory Privacy Preserving Algorithms In Mobile Social Network Services

With the popularization of smart terminals and the development of location technology,location based services have been widely applied,and a new social network,the mobile social network,has been generated.Users on mobile social networks can share and use LBS applications anytime and anywhere,while they also pose a risk of privacy breaches.The trajectory privacy reflects the life habits,religious beliefs,home work addresses and other sensitive information of mobile users over a long-term,and is closely related to the personal safety of mobile users.Therefore,ensuring the trajectory privacy of mobile users is one of the important factors to drive the development of mobile Internet.In recent years,trajectory privacy has been extensively researched by scholars at home and abroad.However,the LBS application scenarios are constantly updated,and the attackers' attack methods are increasingly sophisticated,trajectory privacy protection faces new challenges.Based on the in-depth analysis of the research results of trajectory privacy protection technology at home and abroad,this thesis starts research work from the trajectory privacy protection in the process of using trajectory data.The main work and innovations of this thesis include:1.For mobile social networks,aiming at the balance between mobile users that need to provide accurate location information to service providers to obtain high-quality services and realize the trajectory privacy protection.A personalized trajectory privacy protection algorithm based on dummy trajectory technique is proposed.Firstly,the algorithm used the Euclidean distance to select the nearest k users as the equivalence class to realize the location privacy protection,and then constructs k trajectories similar to the mobile user trajectory to realize the trajectory privacy protection.Then,a threat model and an assessment framework were introduced.Finally,the simulation experiment is carried out,and the effectiveness of the scheme is verified from two aspects: the similarity of the constructed dummy trajectory and the real trajectory,and the level of privacy protection.2.For the reasoning attacks based on association analysis to obtain the privacy information of mobile users,this thesis proposes an optimal deployment strategy based on mix-zone nodes.In a large-scale urban area,a limited number of mix-zone nodes are deployed to fully meet the needs of mobile user trajectory privacy protection and high-quality LBS services,eventually converting the mix-zone deployment scenario to a siting problem.The algorithm establishes an integer programming model by setting the number of deployments,intersection traffic,and information entropy,and introduces an algorithm for selecting mix-zone nodes.Then,the threat model and evaluation framework are introduced.Finally,simulation experiments are carried out to verify the effectiveness of the scheme from two aspects: accessibility and anti-correlation analysis.