LoRaWAN – Based Indoor Localization for Mobile Nodes Using Machine Learning Technique

With the increasing demand for location-based services driven by the expansion of Internet of Things (IoT) devices, traditional localization methods like GPS face limitations in indoor environments due to signal degradation. An approach using LoRaWAN communication and machine learning techniques is proposed to accurately estimate the position of a mobile robot based on Received Signal Strength Indicator (RSSI) signals. A radio map was constructed using data collected from three static LoRa anchors, and a supervised machine learning method based on the XGBoost algorithm was used to predict the robot's location. The experimental setup, which includes data collection, model training, and testing, demonstrates the effectiveness of this approach. The model's performance was evaluated using key metrics such as Mean Absolute Error (MAE) of 0.24 meters, and Average Error (AE) of 0.32 meters, indicating that the proposed system is highly accurate and suitable for real-time mobile node localization in indoor environments