Performance Analysis of Outdoor 5G Fingerprinting in Urban Environments

This study analyzed the performance of 5G fingerprinting in commercial base stations in a dynamic scenario involving an urban environment. A synchronization signal block was used to identify the physical cell identity and synchronization signalreference signal received power of multiple base stations, which were recorded in a fingerprinting database. A universal software radio peripheral (USRP) was subsequently used to collect 5G signals and update data at a frequency of 50 Hz, which significantly shortened the database construction time relative to the time required for data updates on mobile phones, which occur every few seconds. In addition, a dynamic database construction approach was adopted. This approach included the output of a high-accuracy GNSS-INS integrated system as a reference point, with 5G signals continuing to be received in motion. Moreover, a weighted Knearest neighbor algorithm was selected as the positioning algorithm. USRP X300 and LP0965 antennas were used in dynamic experiments to receive 5G signals, and NovAtel PwrPak7 and LN200C were used to obtain ground-truth positions. The results indicated a root-mean-squareerror of 12.02 and 8.67 m in the east and north directions, respectively, with a horizonal error of 16.94 m.