Collaborative location estimation for confined spaces using magnetic field and inverse beacon positioning

A considerable number of indoor positioning systems have been proposed for large-scale environments that extend over several meters. However, there has been less focus on designing an indoor localization system for confined environments where the requirements of reliability and precision are high. The approach discussed in this paper employs a hybrid technique where Received Signal Strength Indicator (RSSI) fingerprinting using beacon and magnetic field sequential data are used in tandem to determine the zone where a person of interest is performing an activity. An offline location signature database is built for each of these methods where the RSSI and the magnetic field sequences are mapped with their respective walking routes or stationary positions within the household. We use the results of RSSI fingerprinting as the basis to narrow down the searching space for the magnetic field vector matching, and later use the Dynamic Time Warping (DTW) algorithm to deduce the similarity between the measured and offline magnetic field sequences. We compare our approach with another variant of DTW known as Derivative Dynamic Time Warping (DDTW) and evaluate the performance of each variant with and without the RSSI fingerprinting stage. To the best of our knowledge this is the first study of its kind that focuses on interpreting fine-grained deviation between positions or routes confined to a small area.