Robust Cooperative Wi-Fi Fingerprint-Based Indoor Localization

Abstract

Wi-Fi fingerprint-based localization has attracted significant research interest recently. Previous works in this area mainly focus on locating an individual user, whereas the additional assistance from peer-to-peer interactions has not been fully exploited. In this paper, we propose a cooperative localization method which not only utilizes the initial results by the fingerprint-based algorithm but also takes into account the physical constraint of pairwise distances to refine the localization estimates for multiple users simultaneously. The experimental results demonstrate that our algorithm is robust against the ranging error and the outdated fingerprint database. With the proposed peer selection scheme, it considerably improves localization accuracy. We further extend our framework to single-user motion tracking and localization based only on access-point-connectivity data.