Radio SLAM-based 5G RSS Ranging in GNSS-Constrained Environments

Fifth-generation (5G) mobile communication system has been widely studied as an alternative or complementary positioning technology to standalone Global Navigation Satellite System (GNSS) positioning. However, present 5G-based positioning approaches typically leverages downlink reference signals, which request extra hardware costs and computational resources. Besides, position estimation derived from measurements including Time of Arrival (TOA), Time Difference of Arrival (TDOA) and Angle of Arrival (AOA) relies heavily on precise time synchronization, making it less reliable in asynchronous networks. To address these limitations, a centroid-distance clustering (CDC)-based stay-point extraction method is proposed to identify stationary segments along the trajectory and mitigate random fluctuations caused by measurement noise. Moreover, a radio simultaneous location and mapping (SLAM)-based estimation method for base station (BS) parameter estimation is developed based on Received Signal Strength (RSS) measurement, enabling adaptive positioning without prior knowledge of BS location, transmit power or antenna gains. Field experiments are conducted using commercial smartphones and operational 5G BSs, with a GNSS receiver providing ground-truth trajectories. The RSS-based ranging is carried out in both static and dynamic scenarios. The results show that the proposed method achieves competitive ranging accuracy of about 20 m, and the CDC-based stay-point extraction further improves the accuracy by approximately 4 m compared with non-clustering approaches.