Research On Location Privacy Protection In Mobile Social Networks

With the speedy development of mobile Internet technology revolution,the number of users under the mobile social network also increases greatly every year.The application service based on mobile location brings great convenience to people's life.The traditional social platform of the computer is gradually changing to the mobile one,which eventually forms a mobile social environment with the integration of social,mobile and location services.Meanwhile,it also produces a lot of track data.Analyzing and mining these trajectories is of great importance to the construction of smart city and Internet plus.However,the release of trajectory data may reveal personal privacy.The development of mobile social network has brought new challenges.The current location privacy protection method can not adapt to the increasingly perfect big data analysis technology.How to effectively protect the trajectory privacy in the mobile social network scene is an important direction of privacy protection at present.As a kind of privacy protection method with strict mathematical definition and theoretical basis,differential privacy technology has been applied in the data release scenarios of relational databases in many industries.However,differential privacy protection for trajectory data is still a new research field,and there are two main problems to be solved: on the one hand,due to the fixed time interval of user trajectory acquisition,there are a large number of dense areas in the original trajectory,so how to preprocess the trajectory data is a big difficulty.On the other hand,in the actual differential privacy protection skill,we achieve differential privacy protection by adding random noise mechanism,which will cause the problem of uncontrollable added noise.We should consider a more reasonable noise allocation mechanism.In order to solve the problem of data reduction and differential privacy budget allocation in trajectory data protection,this paper proposes a new differential privacy trajectory publishing method based on prefix tree.First of all,the concept of user behavior intensive area in trajectory is proposed.All the location points in the behavior intensive area are replaced in the form of center points.On the premise of keeping the original trajectory data basically,a large number of locations with similar physical positions in the trajectory are simplified.Then,using the fitness of the location node in the prefix tree,combined with the inverse distribution of the Laplace cumulative distribution function to add some noise,to build the noise prefix tree.Finally,through the subsequent processing of the consistency process of the noise prefix tree,the social mobile trajectory data with high availability under the condition of differential privacy is finally released.Through the real track data set,the method put forward in this paper is compared with existing methods,and evaluated from the standard of average error of count query.The experiment shows that the differential privacy track protection method in this paper effectively improves the usability of published track data under the condition of ensuring the user's track privacy information.