Adaptive room-level localization system with crowd-sourced WiFi data

Abstract

WiFi received signal strength (RSS) fingerprinting is a promising method for indoor localization but it faces the challenges of a laborious and time-consuming off-line survey process for radio map fingerprints formation, and of variability in the WiFi coverage over time. To address these challenges, recently researchers have begun to consider the concept of crowd-sourcing and automatic floor map and radio map construction. In this paper, we propose an adaptive room-level localization system (ARLS) which focuses on using massive crowd-sourced WiFi RSS data for recognizing different rooms that exist in the coverage area, for determining their locations on the floor map, and for establishing the radio signatures inside the rooms. For the system to accomplish these tasks, all it takes in the off-line stage is for a surveyor to walk randomly through the coverage area to collect two reference RSS traces, and a corridor-level floor map and initial radio map along with points of interest (POIs) will be built by the system automatically. In the on-line stage, unlabeled crowd-sourced user data is gathered to extract room-level information to the map and conduct continuing refining and updating. Our results show that rooms can be effectively recognized by their RSS fingerprints, and that rooms can be localized on the floor map by analyzing RSS traces as users enter and leave a room. The RSS fingerprints of rooms can also be adaptively updated using crowd-sourced user data.