A multisource and multivariate dataset for indoor localization methods based on WLAN and geo-magnetic field fingerprinting

Abstract

Indoor localization is a key topic for the Ambient Intelligence (AmI) research community. In this scenarios, recent advancements in wearable technologies, particularly smartwatches with built-in sensors, and personal devices, such as smartphones, are being seen as the breakthrough for making concrete the envisioned Smart Environment (SE) paradigm. In particular, scenarios devoted to indoor localization represent a key challenge to be addressed. Many works try to solve the indoor localization issue, but the lack of a common dataset or frameworks to compare and evaluate solutions represent a big barrier to be overcome in the field. The unavailability and uncertainty of public datasets hinders the possibility to compare different indoor localization algorithms. This constitutes the main motivation of the proposed dataset described herein. We collected Wi-Fi and geo-magnetic field fingerprints, together with inertial sensor data during two campaigns performed in the same environment. Retrieving sincronized data from a smartwatch and a smartphone worn by users at the purpose of create and present a public available dataset is the goal of this work.