Research On Trajectory Privacy Preservation Technology In Location Based Service

In recent years,with the popularization of smart terminals such as mobile phones and the development of mobile computing technology,Location Based Service(LBS)has become more and more popular.While it brings convenience to people,it also increases privacy leakage.An attacker can easily extract the user's home address,health status,income and other information from the location information.At present,there are two main ways to leak the user's location privacy:(1)Some organizations publish the user's location data for the purposes of scientific research.After the data is obtained by an attacker,it uses big data and other related technologies to mine the user's privacy information.(2)When users receive location services,they are attacked by attackers,resulting in privacy leaks.Therefore,it is of great significance to carry out related research on user privacy protection technology in LBS.Users' privacy in LBS mainly includes location privacy and trajectory privacy.Compared with the latter,the former be studied earlier,more mature,and less research difficulty(the locations in the trajectory have strong spatial and temporal correlations,and more factors need to be considered).Therefore,the research focus of this paper is to effectively protect the privacy of user trajectories.Specifically,the research contents and contributions of this paper are as follows:(1)Differentially Private Trajectory Synthesis Using Hierarchical Reference Systems based on trajectory generalization is proposed to solve the problem of user privacy leakage during the trajectory publish process.Unlike previous work,this paper first uses a Hilbert curve-based space partitioning technique Generalize the original trajectory to reduce the sparseness of the trajectory distribution.Then use hierarchical Reference Systems(rather than a single)to capture the motion characteristics of users at different speeds.Then a new pruning strategy based on binomial distribution is used to prune the prefix count tree.Finally,a direction-weighted sampling algorithm is proposed to restore the trajectory's movement direction and improve the practicability of the data.The experimental results show that the system can ensure the practicability of the data while protecting the privacy of the user's trajectory.(2)Synthesizing Privacy-Preserving Trajectories with Semantic-aware Dummies for Location-based Services is proposed to protect the trajectory privacy of users in LBS.Hiding the real trajectory in the synthetic trajectory is a common way to achieve privacy protection,which makes it difficult for attackers to distinguish between them.The difficulty lies in how to ensure that the synthetic trajectory and the real trajectory are plausible,so as to confuse the attacker.Most previous studies ignored factors such as semantics,granularity,and direction.Attackers can easily distinguish the true trajectory on the premise of mastering certain background knowledge.This paper adopts a newly proposed location-based semantic clustering algorithm to cluster position points in the trajectory data set,and designs corresponding indicators to measure the semantic similarity,granularity similarity,and direction similarity between synthetic trajectory and real trajectory.It also uses a phased selection method to avoid a lot of unnecessary calculations and improve the efficiency of the system.Since the false locations on the synthetic trajectory are in the data set,the synthetic trajectory data set and the original trajectory data set have similar statistical characteristics,so the synthetic trajectory obtained by the algorithm can be published instead of the real trajectory,reducing the risk of user privacy leakage.The experimental results show that the synthetic trajectory obtained by this algorithm is highly similar to the real trajectory in semantics,granularity and direction.