TY  - JOUR
T1  - Indoor Multi-Floor Localization Based on Location Calibration Combined with Magnetic Maps and Deep Learning Models
AU  - Kim, Jin-Woo 
AU  - Shin, Yoan 
JO  - The Journal of Korean Institute of Communications and Information Sciences
PY  - 2024
DA  - 2024/1/1
DO  - 10.7840/kics.2024.49.10.1354
KW  - Pedestrian dead reckoning
KW  - Smartphone-based multi-floor indoor localization
KW  - Deep learning
KW  - Magnetic field map
KW  - Real-time location tracking
AB  - In this paper, we propose a method to address the issues in smartphone-based pedestrian dead reckoning (PDR) by combining magnetic maps and deep learning models to calibrate the current location using calibration nodes when floor transitions are detected, thereby improving localization accuracy. In addition, we verified the localization performance of the proposed method by connecting the server and smartphone via socket communication, allowing real-time localization tracking.