Research And Application Of Location Privacy-preserving Algorithms In Vehicular Networks

Thanks to the rapid development of vehicular networks,people can enjoy the convenience and fun brought by a variety of location-based services(LBS)while driving.When drivers and passengers use these services,they often need to provide the location of the vehicle,which implies personal privacy information,to the corresponding LBS providers,leading to a great threat to the privacy security of users.Given that most LBSs in vehicular networks require users to send service requests continuously,and these continuous LBSs contain more privacy information than snapshot LBSs,this dissertation focuses on dealing with the location privacy leakage problem when users of vehicular networks using continuous LBSs.More details are shown as following three aspects:(1)Most existing schemes do not fully consider the adversary attack model,and assume that all vehicles are using the same LBS.Regarding this,this dissertation proposes a novel collaborative inter-vehicle location privacy-preserving algorithm by analyzing the classical adversary attack model.For each vehicle,Kalman filtering algorithm is applied to predict the location.A suitable partner vehicle is selected to help sending false LBS service requests for a continuous period.Thus,obfuscated trajectories can be generated and location privacy can be protected.Experimental results show that the proposed algorithm can effectively improve the location privacy strength of vehicles and reduce the communication cost.(2)To address the problem that collaborative obfuscated trajectories cannot be generated in low traffic density scenarios,which leads to the inability to protect users’ location privacy,this dissertation designs a personalized location privacy protection scheme.This scheme consists of a false trajectory generation algorithm based on location offset and a custom pseudonym update strategy based on multiple linear regression.The experimental results show that the scheme can not only provide users with high-intensity location privacy protection but also reduce the pseudonym consumption of vehicles under the loss of certain service quality.(3)Combining the respective advantages of collaborative obfuscated trajectory generation algorithm and false trajectory generation algorithm based on location offset,this dissertation proposes a hybrid trajectory obfuscation scheme to cope with the ever-changing traffic density in real scenarios.The performance of the combined scheme is investigated using real datasets in SUMO simulator.The results show that the proposed scheme can provide effective location privacy protection while reducing the quality of service loss.