Research On Location Privacy Metric Based On A Common Framework LPPM

With the popularity of location-based services, and the widespread use of mobile social application based on location services, more and more mobile users begin to worry about their location privacy is exposed. A variety of location privacy metrics have also been proposed, but these metrics have been proposed to be based on a different location privacy protection mechanisms, all with some of the shortcomings and deficiencies, and most of the existing measurement methods ignore the attacker’s background knowledge. Therefore, on this issue, we analyze a variety of location privacy metrics, and propose a new measurement method that is called BK-LPPM(Background Knowledge-Location Privacy Protection Mechanisms) base on a common framework LPPM(Location Privacy Protection Mechanisms), and inclusion the attacker’s background knowledge.The LPPM framework introduced the various components of location privacy according to the flow of information from the users to the adversary. The maximization of the user’s location privacy is defined as the adversary cannot correctly link their location and identity over time. This allows us to accurately determine the location privacy, and a detailed description of its associated structures, and different location privacy preserving mechanisms with respect to different attacks.In order to correctly measure the effectiveness of a variety of privacy protection mechanisms under different attack models, we analyze the various factors influencing the user’s location privacy in LBS system and the relationship between them, and taking into account the attacker’s background knowledge, then propose a new location privacy metrics that is called BK-LPPM which compose of the attacker’s background knowledge, and base on a general location privacy protection framework LPPM. In this measurement method, we quantify the mobile user’s location privacy as the accessibility between attacker observed generalization events and identification information of mobile users. The main features of this method is that the use of conditional probability and correlation vector to describe and quantify the attacker’s background knowledge and its inferential capability, and put the attacker’s background knowledge into a unified location privacy framework, we can more effectively measure location privacy in a variety of location privacy protection technologies, and increasing the portability of location privacy metrics.Finally, the simulation experiment results show that this method is effective.When the attacker have background knowledge, k-anonymous evaluation metrics can not correctly measure the user’s location privacy. However, this article presents a new location privacy metrics that is called BK-LPPM, and it can accurately measure user’s location privacy with attacker’s background knowledge. We propose a entropy anonymous measure evaluation can correctly reflect the location privacy protection mechanism provide to the level of user privacy, and it helps the mobile user to balance the relationship between location privacy requirements and location service quality.