Research On The Method Of Moving Track To Deanonymization

In contemporary society,with the development of positioning technology,a large number of movement trajectories will follow.While technology enriches people's lives,personal privacy is also under great potential threat.Anonymized technology has been used to encrypt user data,and the opposite is deanonymization attack method.Based on the structural features of the trajectory and the correlation degree of the discrete distance attributes between the trajectories,this paper proposes a moving trajectory deanonymization attack method.Firstly,a noise correction trajectory smoothing method is proposed for the noise anomaly data of the moving trajectory.In order to obtain a more realistic and effective deanonymization result,a multi-state dynamic weighted trajectory correction is implemented based on Kalman filtering to realize data preprocessing of the moving trajectory.Secondly,the SA structure attack model is proposed.The SA attack model measures the correlation between the two trajectories based on the similarity of the cluster structure between the trajectories.The RA distance attack model is proposed.The RA attack model measures two based on the degree of dispersion between the trajectories.The correlation between the trajectories is proposed.The TSDM deanonymization attack model combining the SA attack model and the RA attack model is proposed,and the correlation between the overall structural features and the local discrete distance is considered.Then,based on SA attack model,RA attack model and TSDM attack model,the cluster structure similarity algorithm between tracks,discrete distance similarity algorithm and multi-view deanonymization method are proposed respectively.The trajectory is clustered,and the cluster is used as the node to construct the trajectory structure feature,and the structural similarity analysis is performed on the trajectory.The trajectory is sliced and sampled to measure the degree of dispersion.Considering the similarity and discrete similarity of the trajectory cluster center structure,the deanonymization result of the moving trajectory is more robust and accurate.Finally,using the GPS movement trajectory data of 100 users in Beijing,the method of this paper is verified by experimental analysis,and the betterdeanonymization result is obtained,which verifies the effectiveness of the proposed deanonymization scheme for mobile trajectory.