A new grid partitioning technology for location privacy protection

Abstract

Nowadays, the location-based service (LBS) has become an essential part of convenient service in people's daily life. However, the untrusted LBS servers can store lots of information about the user, such as the user's identity, location, and destination. Then the information can be used as background knowledge and combined with the query frequency of the user to launch the inference attack to obtain user's privacy. In most of the existing schemes, the author considers the algorithm of virtual location selection from the historical location of the user. However, the LBS server can infer the user's location information on the historical data that has been counted for a long time. In order to ensure that the users' historical query data and query frequency will not be obtained by the attacker, we propose a privacy protection algorithm based on grid expansion. With the help of third-party agents, the combination of cooperative users and pseudonyms can resist the privacy disclosure caused by users requesting services during the mobile process.Extensive simulation experiments have been carried out on Gowalla dataset to evaluate the efficiency of the proposed algorithm. By comparing with other existing methods, the experimental verify the effectiveness of our algorithm in privacy protection.