Research And System Construction On Privacy Protection Mechanism For Big Trajectory Data

With the popularity of mobile devices and location services,Big Trajectory Data has become an important research area.Big Trajectory Data contains a large amount of personal location information,the disclosure of which may pose a significant threat to personal privacy and security.Therefore,how to protect the privacy of Big Trajectory Data has become an important issue.How can trajectory big data still provide effective support for various applications under the premise of privacy protection.At present,one of the important topics of trajectory data privacy protection is to ensure the integrity,accuracy and availability of data while protecting privacy.Nowadays,most of the trajectory data privacy protection technologies are based on k-anonymity model,differential privacy model or machine learning model,but there are obvious shortcomings in the degree of privacy protection,data availability or protection efficiency.Therefore,we urgently need a trajectory data protection scheme that can meet the requirements of protecting users’ privacy and has high efficiency and high data availability.The specific research work of the thesis mainly includes the following three aspects:(1)With the development of 5G and Mobile Edge Computing technology,the user’s movement trajectory only needs to be uploaded to the MEC server in the area to which it belongs,which provides a new way for trajectory data collection.This thesis proposes a trajectory data collection model based on the 5G mobile edge computing,and uses the service characteristics of the MEC server to propose a differentially private protection of trajectory data based on 5G mobile edge computing(5GMECDP).5GMEC-DP first runs the longitude and latitude coordinate differential privacy protection algorithm on the mobile client,and then runs the MEC server truncation mechanism on the MEC server,and judges the coordinate information after adding the plane Laplace noise.If the encrypted location information does not belong to the service range of the current MEC server,it will be projected to the service edge of the MEC server,thereby limiting the amount of noise of the GeoIndistinguishability algorithm,making it suitable for the protection of trajectory data,and it also proves that the 5GMEC-DP algorithm still satisfies the definition of ε-Geo-Indistinguishability.Finally,experiments on the real data set prove that 5GMEC-DP can maintain the high practicability of the data when the degree of privacy protection is high.Finally,it is experimentally demonstrated on real datasets that 5GMEC-DP can improve the performance of data utility metric AQ by 64%～81%and trajectory dataset usability metric HD by 64%～82% when the privacy protection is high.(2)In order to overcome the problems in the current works such as insufficient accuracy,insufficient efficiency and insufficient protection of trajectory data privacy in trajectory data publishing methods,this thesis proposes a trajectory data differential privacy publishing mechanism LGAN-DP based on LSTM-GAN,which uses a long short-term memory network and a generative adversarial network to generate synthetic trajectories,and designs a trajectory loss function to judge the trajectory similarity loss of the synthetic trajectories trained by the trajectory data deep learning model proposed in this thesis,and finally processes the result set of trajectory data publishing by using differential privacy.The experimental results show that LGAN-DP can better guarantee the balance between the privacy and availability of trajectory data on the MI metric.And the availability of trajectory data metric HD is improved by 40%～50% as compared with the existing methods,with less time complexity.(3)A trajectory big data privacy protection and publishing system get constructed based on above technologies,namely trajectory differential privacy protection mechanism based on 5G mobile edge computing and trajectory data differential privacy publishing mechanism based on LSTM-GAN.The system is designed with visualization module,data input module,route planning module,trajectory data encryption module,cache module,and data retrieval module.It implements prediction function using a pre-trained trajectory prediction model,shows and publishes generalized trajectories generated by differential privacy publishing mechanism,encrypts location coordinates using differential privacy for data display and error presentation.The system is based on web technology and realizes the privacy protection and publishing of trajectory big data.