GPS inter-frequency clock bias estimation for both uncombined and ionospheric-free combined triple-frequency precise point positioning

摘要

The time-varying biases within carrier phase observations will be integrated with satellite clock offset parameters in the precise clock estimation. The inconsistency among signal-dependent phase biases within a satellite results in the inadequacy of the current L1/L2 ionospheric-free (IF) satellite clock products for the GPS precise point positioning (PPP) involving L5 signal. The inter-frequency clock bias (IFCB) estimation approaches for triple-frequency PPP based on either uncombined (UC) observations or IF combined observations within a single arbitrary combination are proposed in this study. The key feature of the IFCB estimation approaches is that we only need to obtain a set of phase-specific IFCB (PIFCB) estimates between the L1/L5 and L1/L2 IF satellite clocks, and then, we can directly convert the obtained L1/L5 IF PIFCBs into L5 UC PIFCBs and L1/L2/L5 IF PIFCBs by multiplying individual constants. The mathematical conversion formula is rigorously derived. The UC and IF triple-frequency PPP models are developed. Datasets from 171 stations with a globally even distribution on seven consecutive days were adopted for analysis. After 24-h observation, the UC and IF triple-frequency PPP without PIFCB corrections can achieve an accuracy of 8, 6 and 13mm, and 8, 5 and 13mm in east, north and up coordinate components, respectively, while the corresponding positioning accuracy of the cases with PIFCB consideration can be improved by 38, 33 and 31%, and 50, 40 and 23% to 5, 4 and 9mm, and 4, 3 and 10mm in the three components, respectively. The corresponding improvement in convergence time is 17, 1 and 22% in the three components in UC model, respectively. Moreover, the phase observation residuals on L5 frequency in UC triple-frequency PPP and of L1/L2/L5 IF combination in IF triple-frequency PPP are reduced by about 4mm after applying PIFCB corrections. The performance improvement in UC triple-frequency PPP over UC dual-frequency PPP is 7, 4 and 2% in terms of convergence time in the three components, respectively. The daily solutions of UC triple-frequency PPP have a comparable positioning accuracy to the UC dual-frequency PPP.