Concurrent Queries in Location Based Services

The location-based services (LBS) are gaining a great importance due to the tremendous development in telecommunications and geographic information systems. In addition, the spread of laptops and mobile devices that can use these services at anytime and anywhere helps the diffusion of this type of application. Nevertheless, there are significant challenges that could impede the use of this type of services. User's Privacy is considered as one of these important challenges that limit the usage of these services because this type of services is user location based. Thereby, different techniques are developed to preserve the user location privacy while s/he uses the LBS. Most of these techniques depend on the Kanonymity [11] of user's location by querying about a spatial region that contains k-1 users. Thus, the adversary cannot detect the user who asked for the service from these k users. However, if the adversary is in possession of some information about the users such as users' profiles and historical queries, s/he can discover in a high percentage the query issuer from the k users in the spatial region. In addition, if several users in a specified spatial area send at the same time their requests, the potential attacks targeting the privacy in the LBS could increase. In this paper, we investigate the impact of users' profiles and previously issued queries by the users in case of concurrent queries on the privacy. We propose an algorithm for predicating the issuer of each query and her/his underlying location in case of concurrent queries based on the users' profiles, historical queries, and the spatial area geographical characteristics. The experiments show that the concurrent queries affect negatively the privacy level in the location based services.