Privacidade de localização geográfica em consultas a serviços públicos web de localização: uma abordagem baseada em médias aleatórias

Abstract

The increasing demand for Location-Based Services (LBSs) has increased the need for solutions that consider the privacy of queries in public LBSs. The currently proposed approaches rely mainly on the construction of an anonymizing spatial region (ASR) based on a rectangle containing a user consultant and at least other K−1 users, thereby setting a K level of anonymity for the query performed. In parallel, P2P infrastructures have been designed to bear the burden of mobility that has characterized much of LBSs users, given the remarkable technological advance that provided the availability of modern resources (such as processing of the GPS signal) on portable devices, such as smartphones. However, the answers obtained from LBSs, considering the context of research conducted, is limited to the indication of points of interest near the consultant user s location, not being able to indicate a route to the destination indicated. This paper aims at presenting the proposal AnoniMobi, whose main objective is to provide routes to points of interest in response to anonymous queries made to public LBSs, ensuring the privacy level requested by the consultant user. The approach separates itself from traditional proposals for formation of spatial region anonymization (ASR) when considering the use of points (random means) that allow you to group coordinates in a non-deterministic way. Additionally, an infrastructure based on CAN P2P architecture is designed to support self-organization needed for structuring and grouping users of the system.