Indoor Geofencing Based on Sensorless Motion Sensing and Fingerprint Self-Updating

TLDR: This paper proposes a hybrid electronic geofence approach that combines self-updating RSS fingerprints based localization and Channel State Information (CSI) motion detection and Multidimensional matching and filtering principle achieves fingerprints self- updating and improves the localization accuracy.

Abstract: Indoor location is definitively a key feature with immense value especial for geofencing. The received signal strength (RSS) fingerprinting based methodology is widely adopted to determine his/her proximity to that particular region. Its dynamic nature and maintain overhead remain a primary challenge. In this paper, we propose a hybrid electronic geofence approach that combines self-updating RSS fingerprints based localization and Channel State Information (CSI) motion detection. Multidimensional matching and filtering principle achieves fingerprints self-updating and improves the localization accuracy. CSI-based speed estimation reduces localization frequency and overhead. Our extensive real-world experiment results show that the proposed indoor geofencing method works well for more than 30 days without manual Wi-Fi fingerprints updating.