Performance analysis of GPS/Galileo/BDS three systems and all-frequency PPP-AR

The accurate estimation of phase biases is an essential prerequisite for achieving precise point positioning ambiguity resolution (PPP-AR). However, the phase bias products provided by the current International GNSS Service (IGS) analysis centers (ACs) are limited in terms of frequency combinations, which restricts PPP-AR users to using observations from corresponding frequencies only. This results in the underutilization of multi-frequency GNSS observations. In order to realize flexible PPP-AR based on arbitrary GNSS frequency selection and observation combinations (referred to as all-frequency PPP-AR), we conducted an analysis using static observation data from over 100 globally distributed IGS stations for a week. We estimated the observable phase biases (OSBs) for all available frequencies and conducted static precise point positioning experiments using the open-source PRIDE PPP-AR software. The results indicate that the average standard deviations of pseudo-range and phase OSB products are 0.25 ns and 0.34 ns, respectively, satisfying the requirements for PPP ambiguity fixing. The average ambiguity fixing rates for Galileo and BDS-3 frequency combinations are 98.25% and 96.74%, and 90.31% and 91.64%, respectively. For BDS-2, the average ambiguity fixing rates are 81.27% and 86.02%. Compared to the reference frequencies, the positioning accuracy decreases by nearly 15% in all three directions for E1/E6 and G1/G5 combinations. The C2/C5 combination performs well in the east (E) and north (N) directions but exhibits a 22.8% decrease in accuracy in the up (U) direction. In the testing of the GPS, Galileo, and BeiDou Navigation Satellite System (BDS), the remaining frequency combinations, including C2/C7, C1/C5, C1/C6, C1/C7, E1/E7, and E1/E8, all exhibited stable positioning accuracy. The positioning accuracy of each frequency combination varied by 5% in the E, 5% in the N, and 15% in the U direction, with no significant deviation compared to the reference frequency. Therefore, full-frequency PPP-AR has achieved a considerable degree of reliability. Full-frequency PPP-AR takes full advantage of the multi-frequency GNSS observations, offering higher practical value compared to using observations from the base frequency alone. Furthermore, full-frequency PPP-AR allows users to select and combine GNSS frequencies according to their specific requirements, thereby better satisfying the needs of various applications.