Research On Trajectory Privacy Protection Method For Internet Of Vehicles Mobile Applications

Vehicle positioning technology has developed vigorously in the past few years.At the same time,the integration of mobile positioning technology for vehicles and self-organizing networks has brought about a thriving Internet of Vehicle technology.The demand for location-based services in the Internet of Vehicles has surged,and the trajectory information generated by vehicle movement can provide support for intelligent driving,traffic management,and data analysis.However,there is a lot of user traj ectory information in the Internet of Vehicles,which leads to the risk of data security and privacy leakage of vehicle trajectory data.It is of great significance to study how to realize the privacy and security of vehicle traj ectories on the premise of ensuring the accuracy of location services in mobile Internet of Vehicles applications.This paper focuses on the privacy protection of Internet of Vehicles location services.The specific research work is as follows:(1)In order to solve the problem of privacy leakage caused by low location anonymous generalization,large time overhead and low privacy in non-interactive location services in the Internet of Vehicles,the traj ectory privacy protection algorithm MPBD-SimHash(Multiparameter blend based on multidimensional attribute matrix calculation is firstly proposed dimension-SimHash).The algorithm selects the candidate groups of collaborative users according to the user distance,and then sorts the users in the candidate groups according to the similarity of attributes to reduce the generalization of location data,quickly obtain highly similar collaborative objects,and realize the protection of privacy information.Then,on the basis of the MPBD-SimHash algorithm,a false trajectory generation algorithm MSPA(MPBD-Simhash blending of pitch and angle algorithm)is proposed to ensure that the generated false vehicle traj ectory and real vehicle trajectory The similarity of the motion patterns is high and the direction of travel is similar,which improves the similarity of the trajectories.Experiments show that the MSPA algorithm can ensure the high similarity between the collaborative vehicle and the generated traj ectory in the anonymous area,which improves the degree of privacy and has a greater advantage in time.(2)In order to solve the problem of low location service utility caused by the clustering of location points sensitive to outlier parameters in the interactive location service in the Internet of Vehicles,a spatiotemporal prediction obfuscation method based on density clustering is proposed to achieve privacy protection of vehicle trajectories.First,a polygon anonymous area generation algorithm based on stagnation point density clustering,Polygon-optics,is used to optimize the parameter values and outliers,which solves the problem of high privacy leakage caused by sharing anonymous areas.Then,a multi-region united trajectory prediction privacy protection algorithm(MUTP)algorithm is proposed,which uses LSTM to predict the next position based on historical information to disturb the spatio-temporal information of the vehicle,protect the real traj ectory of the vehicle,and prevent A timing attack due to shared information.Experiments show that the MUTP algorithm has a good prediction effect,and achieves a good track privacy protection.(3)In order to solve the problems of low privacy protection effect and poor data availability caused by the addition of uncorrelated noise in the release of vehicular vehicle trajectory data,a personalized differential privacy protection method based on multiple correlation function constraints is proposed.First,the historical best match length algorithm HBML(Historical best match length)is used to optimize the calculation of the privacy value of the location point,and the problem of unreasonable privacy value generation is solved.Second,the personalized differential privacy protection algorithm NAPDP(Noise adding algorithm for Personalized differential privacy)Add personalized noise that matches the characteristics of the trajectory data;Then,a TPMCP(Trajectory protection method of multiple correlation function with parameter)fusion track distance fusion algorithm is proposed to ensure that the generated false trajectory and real trajectory have The same movement mode realizes trajectory data protection.Experiments show that the algorithm can better protect the trajectory privacy of vehicles.