Research On Trajectory Privacy Protection Method Based On Movement Trend

With the development of positioning technology,wireless communication technology and smart mobile devices,location-based services(LBS)have been widely used in all aspects of life.The LBS provider(LSP)can provide location-based services based on the location information that users get from their smartphone.This has facilitated people's lives and strengthened their dependence on LBS.However,users do not know how the LSP handles their location information,which raises people's concerns.In continuous query,the LSP can track the user's location and retrieve the user's trajectory,leading to the leakage of trajectory privacy and threatening the privacy of the user.Many methods have been proposed in continuous queries to protect the user's trajectory privacy.At present,most methods rely on trusted third party(TTP)and utilize trajectory k-anonymity technology to protect the privacy of users.However,the trajectory k-anonymity technology requires that k users always remain in the same cloaked set during the continuous query,which will cause the cloaking area of the user to be too large and reduce the service quality.Moreover,TTP has the risk of privacy leakage.Once the trusted third party is attacked,it will lead to more serious privacy leakage problems.Aiming at the two problems mentioned above,this paper considers the user's moving trend,and proposes two methods to protect the user's trajectory privacy and improve service quality.The main work is as follows:(1)Aiming at the problems of large cloaking area and low service quality caused by traditional trajectory k-anonymity technology,a trajectory privacy protection method based on location similarity of query destination was proposed to protect the trajectory privacy of users in continuous queries.First,the concept of "Location Similarity Of Query Destination(LS-QD)" is defined to reflect the user's global moving trend.Then,a quadtree model is proposed.By improving the grid-based pyramid data structure in the Casper method,users in the cell are dynamically tracked to achieve rapid retrieval of users who are not yet anonymous.Finally,based on LS-QD,a novel trajectory k-anonymity method is proposed,that is,users who satisfy the characteristics of LS-QD are selected to implement trajectory k-anonymity.Security analysis proves that the method can resist multiple trajectory attacks.Experimental results show that this method can significantly reduce the cloaking area and cloaking time of users,and can improve the service quality.(2)Aiming at the problem of privacy leakage after a trusted third party is attacked,a trajectory privacy protection method based on angle-obfuscated is proposed.First,an angle-obfuscated method is proposed.All query angles located in the same angle interval are replaced with a specific angle and this angle is used as the user's moving trend to construct a trajectory k-anonymity set.Then,a region-adaptive pseudonym changing strategy is proposed.When the cloaking area is too large,the pseudonym is changed to improve the service quality of LBS.By combining dynamic pseudonym technology with the two methods proposed above,the trajectory is cut off from the user's identity,and the long trajectory is divided into mult iple shorter trajectories,which makes it difficult to reconstruct the trajectory and improves privacy level.In this solution,the anonymous server cannot obtain the user's complete trajectory,and cannot associate the trajectory with the user's identity,thereby achieving the purpose of protecting the user's trajectory on the anonymous server side.Security analysis shows that the scheme can resist the strong adversary attack model and the weak adversary attack model,and effectively protect the user's trajectory and query content.Experimental results show that this method can improve the service quality while protecting the privacy of the trajectory,and has good performance.