Trajectory Privacy Protection Research Towards Mobile Location Service

With the development of 5G and the upgrade of mobile smart devices,location-based service(LBS)has been applied in various fields such as transportation,food delivery and logistics tracking.Although LBS has brought new lifestyles to people,the widespread use of LBS can generate large amounts of personal data,including personal trajectory data.The trajectory data includes the user's home address,work unit and other sensitive information.If the information is leaked,the user's privacy security will be threatened.Therefore,how to ensure the privacy of users' trajectories while using LBS applications has become the focus of people's attention.Firstly,this thesis introduces the current research status of trajectory privacy protection at home and abroad.Then the system architecture and related technologies are deeply studied,and their advantages and disadvantages are analyzed.Secondly,according to the two application scenarios of trajectory privacy protection(trajectory privacy protection in data released and real-time trajectory privacy protection in LBS),this thesis proposes corresponding privacy protection schemes.Finally,through experimental comparison and analysis,this thesis proves that the two proposed schemes can effectively protect trajectory privacy and maintain a certain quality of service.The main research contents of this thesis are as follows:(1)In order to protect trajectory privacy in data released and maintain data availability,this thesis proposes a trajectory privacy protection scheme based on location sensitivity division.The location trajectory and check-in frequency are firstly used to set thresholds in this scheme,and then the location sensitivity levels are classified.Secondly,considering that adding uniform noise to the location causes excessive error and cannot satisfy the user's individual needs,a corresponding privacy budget is allocated according to the sensitivity level in this thesis,and Laplace noise which meets differential privacy is added.Finally,through experimental analysis,this scheme can effectively protect trajectory privacy in data released and reduce the interference of differential privacy on service quality.(2)To improve the success rate of dummy trajectory schemes in LBS,a location obfuscation trajectory privacy protection scheme based on collaborative filtering(TPCF)is proposed in this thesis.TPCF uses a trusted third-party anonymous server.Firstly,the user asks for privacy protection while sending a request,and the anonymous server then recommends several candidate locations based on demand.Secondly,dummy locations among candidate locations which satisfy spatial accessibility,temporal accessibility,directional consistency,and location dispersion are selected.Finally,the user's real location and dummy locations are sent to the location service provider(LSP)at the same time.Through theoretical analysis and experiments,not only the risk of users' real location exposure is reduced,but also real-time service requirement is met by TPCF.