Indoor Positioning Using WLAN Signal Fingerprint Matching and Path Evaluation with Retroactive Adjustment on Mobile Devices

Abstract

The availability of GNSS has spurred a plethora of location-aware services. However, GNSS are only reliably available outdoors, leaving a large gap for a solution in indoor environments where people spend most of their time. The ubiquity of mobile devices, and the many information sources available on them, provide many avenues for a potential indoor positioning solution. A novel indoor positioning system is presented herein, using WLAN Signal Fingerprint Matching (WSFM) and a Path Evaluation and Retroactive Adjustment (PERA) module. The PERA module aims to improve the positioning accuracy by fusing the results obtained by WSFM with a multi-scale movement regularity evaluation. Using only WSFM, the implemented end-to-end positioning system yields room level positioning accuracy (less than 3-5 metres of error) 90% of the time across various environments. Employing the PERA module reduces positioning error to less than 2-3 metres 95% of the time across all testing settings.