Research On The Privacy Protection Method Of Trajectory Data Based On Differential Privacy

In recent years,with the rapid development of mobile devices and computer network systems,mobile consciousness of service systems are released to a third party service providers and a significant increase in the amount of personal data,and the increase of the amount of data,in turn,leads to data privacy leakage hidden trouble.For example,trajectory information released without protection may expose the user's physical condition,work and rest habits,etc.More seriously,location information leakage may also lead to security problems,which may be used by criminals and threaten the user's personal safety.Therefore,it was of great significance to study the privacy protection of trajectory data.At present,there are many methods to protect the privacy of trajectory data.These methods are time-consuming to implement,unable to process large-scale trajectory data,and have poor availability of protected data.In view of the above problems,this paper proposes a method of releasing trajectory data based on differential privacy mechanism,and constructs and publishes processed trajectory data to prevent privacy leakage.As a first step,in the phases of data pretreatment,this paper proposes a high efficiency and fast segmentation approach,after the track data segmentation,will visit the frequent users track access point form polygon,polygon centroid calculation,using polygon center instead of the point being protected,to reconstruct for trajectory,so that can handle the massive trajectory data set;Second,the noise track prefix tree structure is used to represent the reconstructed track data set.Laplace noise is added to the nodes of the noise track prefix tree based on the adaptive privacy budget allocation method,and then published after consistency processing.Through experiments,the internal parameters of the proposed method in this paper were determined,and then compared with Prefix and n-gram,which were the proposed trajectory privacy protection algorithm.Experiments show that the algorithm in this paper has good privacy protection effect and high data availability in trajectory data release.