Research On Trajectory Information Protection Mechanism And Its Application Based On Differentiated Privacy

With the increasing popularity of mobile smart terminals and the corresponding rapid development of positioning technologies,location-based services have been widely integrated into people's daily lives.Such applications use the location service to obtain the location information of the user and provide relevant third party with the corresponding service(such as querying the weather).According to research statistics,there are about 80% of the applications that people use in their everyday lives,and the services related to location information.Location-based services can greatly facilitate people's lives and improve their living efficiency,but the large amount of user location information that comes with them contains a lot of sensitive information,which may be produced by criminals,causing users' private information to be leaked.The private information of the home address,work location,sexual orientation and political views may even threaten the personal safety of the user.Therefore,while ensuring people's convenience brought by location services,protecting the user's location privacy is an urgent problem to be resolved.Spatial indistinguishable is a location privacy protection mechanism based on differential privacy.That is,the probability that the input ends at the same point is similar when the input is two points that are not divided by a given distance.In this way,the attacker cannot infer the user's real location from a bunch of adjacent locations,thus protecting the user's location privacy.However,under the premise of guaranteeing privacy,users always want the quality loss of the accessed service to be minimal.In other words,how to minimize the loss of service quality of users under the premise of protecting user privacy.The private goal of location privacy is to try to confuse the user's real location or trajectory,and the privacy goal cannot cover the combination of spatial and temporal information(that is,spatiotemporal events).Although the private model of a single location or trajectory protects the user's location privacy,it does not provide a certain level of privacy protection for spatiotemporal events.Firstly,being aimed at minimizing the loss of quality of service,this thesis proposes a Laplace mechanism for the post-map plane based on the spatially indistinguishable privacy protection model,and improves the service by finding the optimal mapping location point that satisfies the condition.Quality loss and demonstrate the private intensity of the mechanism.Secondly,for the privacy protection of space-time events,this thesis extends the privacy objective of location privacy including only spatial information to cover the combination of spatial and temporal information,and defines the privacy of ?-space-time events,thus giving a privacy protection for time and space events.The mechanism converts the location privacy protection mechanism into a mechanism for privacy protection of spatiotemporal events.Finally,experiments on real data sets to verify the usability of the mechanisms presented in this thesis.