Research On Location Privacy Protection Of Internet Of Vehicles Based On Differential Privacy

The Internet of Vehicles is the product of the close integration of wireless communication technologies and automobiles.It is regarded as the key to intelligent transportation systems and the protection of transportation convenience and safety,and has been widely concerned by people in the industry.Because Internet of Vehicles has a large-scale and fast-changing topology,users are likely to expose their true location when interacting with the location service provider,resulting in an increased probability of the vehicles privacy information being stolen,which makes the Internet of Vehicles facing security and privacy threats.Further,Location privacy protection is one of the contents of Internet of Vehicles privacy research,Which arms to cut off the connection between location information and users identity privacy.Therefore,effective location privacy protection has important research significance.This thesis mainly focuses on research and analysis of location privacy protection,and the work is summarized as follows:1.Users have different privacy requirements for different location,the stronger the privacy protection,the lower quality of location services.Since the differential privacy algorithms cannot balance the conflict between the accuracy of location-based services and the strength of location privacy protection,a differential privacy location protection algorithm based on privacy classification is proposed.Firstly,measure whether the measurement sensor location error matches user's privacy needs,if meets,then directly send measured location to the server;if not,then classify user's real location according to users' custom sensitive keyword through decision tree model.Secondly,the measured location and privacy level are used as the input of the differential privacy gaussian mechanism algorithm,and obfuscation location that meets privacy needs is output.Finally,comparing with geographic indistinguishability algorithms and clustering indistinguishable algorithms by the Geolife dataset,the privacy protection strength of the algorithm proposed in this thesis is relatively high.The simulation results show that the differential privacy algorithm based on privacy classification not only strengthens user privacy protection but also guarantees the accuracy of location-based services.2.In view of the fact that the existing location privacy protection mechanism cannot protect the location information of users during spatiotemporal activities,and publish information to the server increase the risk of privacy leakage,the privacy protection algorithm for spatiotemporal events based on differential privacy is studied.First,the spatiotemporal event is first taken as a privacy goal and formalized as a Boolean expression of time and space dimensions.Secondly,define the privacy of spatiotemporal events through differential privacy,give a spatiotemporal events privacy protection model,and interfere with the real location as a obfuscation location that meets the privacy needs of users.Finally,the data set verification shows that the algorithm meets the privacy requirements of spatiotemporal events.Simulation results verify the availability of privacy protection algorithms for spatiotemporal events based on differential privacy.