Anonymization in Proactive Location Based Community Services

Location information of mobile users is very sensitive with regard to privacy. This is especially true for proactive location based services (LBSs) where users are continuously tracked. Effective privacy enhancing technologies for these services are still an open field of research. In this paper, we focus on proactive location based community services (PLBCSs), such as mobile gaming or buddy tracking. Instead of relating the spatial position of users to fixed real-world entities, like e.g. tourist sights in a navigation service, in PBLCSs, the focus is on a user’s spatial relation to other users or virtual entities. In the paper, we discuss why existing privacy mechanism for LBSs are not well suited for PLBCSs. We also present initial ideas for a novel anonymization technique, where location coordinates are obfuscated by a key shared among community members before they are emitted to a PLBCS. Like this, statistical attacks on pseudonym-associated can be avoided while calculations of spatial relations between the entities can be done with unchanged accuracy. 1. Proactive Location Based Community Services (PLBCSs) and Privacy In ubiquitous computing environments the need for privacy amplifies, as ubiquity of computing leads to reduced user control over collected data [Lang01]. For proactive LBSs this is especially true. Although users agree to be tracked in general, location information is collected continuously and therefore without explicit user knowledge so that events relevant for the LBS, such as e.g. a target reaching a point of interest (PoI) can be recognized. In the following, we would like to focus on proactive Location Based Community Services (PLBCS) like buddy tracking [AEM+04], where community members are notified when their mutual distance falls below a pre-defined threshold, or mobile gaming, where, in contrast to e.g. navigation services that relate user positions with real-world entities, the positions of users are only related with the one of other players or virtual entities. Generally, we identify the following requirements for the PLBCSs considered: • The precision of collected position information must be relatively high, service latency low. • Application areas can include the living-area or workplace of a user. ∗Mobile and Distributed Systems Group Institute for Informatics Ludwig-Maximilian University Munich, Germany email: [georg.treu|axel.kuepper]@ifi.lmu.de, [email protected]