Indoor Localization Method Based on WOA Algorithm Optimization of Xgboost

Due to the complex and dynamic nature of indoor environments, satellite signals are severely obstructed by indoor structures when propagating indoors. Therefore, indoor positioning technologies are needed to replace satellite signals for obtaining location information. With the rapid advancements in communication technology and artificial intelligence, AI-based indoor positioning techniques can effectively achieve indoor localization. This paper proposes a machine learning-based indoor positioning technique that utilizes the WOA-optimized XGBoost algorithm to predict the target location in a WiFi-based indoor positioning system. For dynamic targets, the Unscented Kalman Filter (UKF) is employed to fuse measurements and achieve precise position estimation.