Fuzzy Logic-Aided Hybrid GNSS/5G Positioning With Extended Kalman Filtering in Urban Canyon Environments

Fifth-generation (5G) networks can be integrated with Global Navigation Satellite Systems (GNSS) to improve the positioning accuracy in urban canyon environments, where GNSS signals are frequently degraded by limited satellite visibility and severe multipath effects. Although 5G positioning reference signals (PRSs) provide higher update rates than GNSS, existing hybrid GNSS-5G approaches often fail to fully exploit this advantage. Moreover, most adaptive fusion methods involve high-computational complexity, limiting their suitability for real-time applications in dynamic urban scenarios. To address these challenges, this work introduces a fuzzy logic-based hybrid GNSS-5G positioning scheme (FL-HPS). In the proposed method, a fuzzy logic controller dynamically selects between a standard extended Kalman filter (SEKF) and an adaptive EKF (AEKF), thereby ensuring both robustness and efficiency in position estimation. Simulation results across pedestrian, vehicular, and uncrewed aerial vehicle (UAV) scenarios demonstrate that FL-HPS consistently outperforms GNSS-only and conventional tightly coupled methods. In the UAV case, for instance, FL-HPS achieves a mean positioning error of 1.04 m and submeter-level accuracy in 48.13% of epochs, compared with only 0.94% for a tightly coupled EKF. Beyond accuracy improvements, FL-HPS delivers substantial computational gains. By operating at the position level and avoiding iterative raw measurement processing, it reduces average computation time to less than 0.00025-s per epoch - over 25 times faster than the tightly coupled approach.