Research On Trajectory Privacy Protection Method Based On Differential Privac

Location privacy protection is a very important research content in the application field of LBS(Location Based Services)in recent years.Traditional privacy protection methods for a single location can not solve the problem of privacy disclosure caused by the spatial relationship between multiple locations.Trajectory data generally contains continuous location points in space,and the connections between locations in each individual trajectory often reflect personal information of different individuals.Therefore,it is particularly important to protect personal privacy information as much as possible in the analysis and application based on trajectory data.At present,most research schemes ignore privacy disclosure caused by semantic attribute inference attacks.This paper proposes the following two trajectory privacy protection schemes for privacy disclosure caused by semantic information that is not considered in trajectory privacy protection.Most trajectory privacy protection methods use the connections between geographical locations to carry out multi-location clustering or center point replacement to protect the trajectory location,but ignore the privacy disclosure caused by the correlation between location points caused by semantic information carried by location points.To solve above problems,this paper proposes a mechanism based on the index difference of privacy trajectory protection methods: first of all,based on the method for some users,the concept of mining area,secondly combining semantic properties and stay within the region points of privacy of spatial attribute scoring function design,use the index mechanism of the difference of privacy for trajectory data privacy protection.The proposed scheme takes into account the correlation between semantic attributes and geospatial attributes,and aims to protect the user’s sensitive location from leakage while keeping the trajectory characteristics basically unchanged.Finally,the simulation experiment shows that the proposed scheme can effectively improve the data availability.When using differential privacy noise mechanism to protect location privacy,the level of privacy protection is related to the size of privacy budget.How to ensure data availability as much as possible under the same level of privacy protection is an important goal of this kind of research.In order to achieve differentiated privacy protection,this paper analyzes the privacy protection levels required by different stay areas based on location semantics,so that areas requiring higher privacy levels can obtain higher privacy protection.In this paper,a personalized trajectory privacy protection method based on semantic privacy degree is proposed.Firstly,the trajectory’s stay region set is obtained by using the stay region algorithm.Secondly,the average semantic degree of all the stay regions to be protected is calculated according to the semantic position sensitivity.Allocate different privacy budgets to each region based on different average semantics.Experimental results show that under different privacy budgets,the proposed scheme can effectively reduce noise,reduce trajectory distortion and improve data availability.