Research On Privacy Protection Technology In Location Service

As the basic services of Mobile Internet, Location Based Services (LBSs) play an increasingly important role in people’s daily life. According to statistics, nearly 80% of the applications related to LBSs. While enjoying LBSs, mobile users need to share position information with location service provider. Mobile users will expose their private location to untrusted location service provider, which will inevitably bring privacy issues. The researchers proposed a number of location privacy-preserving methods by using technologies of location k-anonymity, pseudonym, location confusion and encryption. However, due to complex scenarios of LBSs, there are still some critical issues that must be resolved, such as how to balance the service quality and the privacy security, how to design location privacy-preserving methods to meet the individual needs of the user, how to accurately analyze background knowledge and reasoning ability of the attacker and how to design a unified framework to measure a variety of location privacy-preserving methods.In this paper, we conducted deep studies on the location privacy-preserving system structures, location privacy-preserving technologies, location privacy attacks and other related technologies about location privacy. And then we proposed a new location privacy-preserving method and a new evaluation model for location privacy-preserving methods.(1)The proposed location privacy-preserving method got rid of the performance bottleneck of traditional central server structure and solved the vulnerability of single point structure by using mobile peer-to-peer structure. Furthermore, in order to ensure the privacy of user in case of the user agent being non-credible in mobile peer-to-peer structure, our proposed method combined the real road network environment to construct anonymous region for users. Finally, the method used the Incremental Nearest Neighbor (INN) query to solve the problem of uncontrollable communication cost in the traditional INN query, which can return almost exact k-nearest neighbor query results.(2)The proposed evaluation model for location privacy-preserving methods analyzed the important elements of location privacy-preserving methods and built a unified framework for the methods. Under this framework, we proposed several evaluation metrics to measure the privacy degree of the location privacy-preserving methods, which considered attackers’ background knowledge and inference skills. The proposed evaluation model solved the problem that location privacy-preserving method lacked objective and comprehensive evaluation metrics.Finally, we analyzed the feasibility and efficiency of the location privacy-preserving method and the evaluation model by conducting experiments.