The Protection Of Check-in Trajectory Privacy In Location-based Mobile Social Network

The location-based social network services application(LBSNS) greatly enriched user social content. But there is also a higher location privacy risks. How to protect the user’s location privacy is the key problem that needed to solve in the process of promoting LBSNS.First, in order to study the location privacy protection issues in the application effective and rational, specific location privacy in LBSNS service at all levels and the role of the various scenarios, to guide how to protect the location privacy in each level in general. On the basis of the reference traditional location privacy protection framework and trajectory privacy framework, according to the characteristics of location privacy protection in LBSNS application, to put forward a unified and comprehensive location privacy protection framework(LPPF). LPPF framework in general can be divided into three parts, namely the client, the location server and application server three modules. Specifies from the location information which is to be adopted and stored, to it is used in different scenarios, until it is shown to the user, users and applications use of location information in different stages. LPPF can provide convenient privacy protection strategy to users, establishing location privacy protection model, providing perfect protection mechanism for the user’s location privacy protection.Second, under the guidance of the principles of privacy protection LPPF architecture, the check-in trajectory of user’s privacy protection in LBSNS applications were studied. Trajectory of privacy protection is a special kind of location privacy protection, but there is lacking of targeted trajectory privacy protection algorithm in LBSNS. In order to solve the problem, the k-anonymous algorithm is proposed for check-in trajectory(KACT) based on decision tree algorithm. Algorithm using the user’s intended sequence to construct prefix tree, adopting the prefix tree representation check-in sequence can save storage space, through the prefix tree pruning, cutting off the branch and leaf nodes which do not meet the parameter k, and use the cut off the branch with pruning prefix tree to reconstruct prefix tree, building k-anonymous prefix tree, achieving check-in sequence k-anonymity, to ensure the safety of users check-in trajectory.Finally, it uses the Gowalla database which contain user’s check-in data as the simulation data in the thesis, using the(k,?)-anonymity trajectory privacy protection algorithm as comparison algorithm, and through the simulation analysis of privacy protection parameter k, the length of the check-in sequence, the longest tolerance ?t. The results showed that under the condition of different parameters, the k-anonymous algorithm for check-in trajectory can guarantee high check-in success rate and a shorter running time. Prove that the k-anonymous algorithm for check-in trajectory has better availability and efficiency in the protection of the privacy trajectory protection.