Analysis of the carrier-phase multipath in GNSS triple-frequency observation combinations

Multi-frequency carrier-phase linear combinations are greatly beneficial to improving the performance of ambiguity resolution (AR), cycle slip correction as well as precise positioning. However, as a limiting factor for further improvements in Global Navigation Satellite System (GNSS) high-precision positioning, the effect of multipath is also accentuated due to the measurement combination. In this paper, with real GPS and BDS triple-frequency observations, we analyse the influence of the carrier-phase multipath in three typical triple-frequency combinations: extra-wide-lane (EWL) combination, ionosphere estimation with ambiguity-corrected EWL/wide-lane (WL) combinations and the geometry-free and ionosphere-free (GIF) combination for narrow-lane (NL) AR. For more intuitive reflection of the influence caused by the carrier-phase multipath, zero-baseline tests with no multipath influence were carried out for comparison. The results from consecutive orbital repeat periods were also used to confirm this influence. Experiments show that for the BDS (1, 4, −5) and GPS (1, −6, 5) EWL combinations, the multipath errors could be combined with the biases of several metres in units of distance and over 0.5 cycles in units of cycles. Therefore in single-epoch EWL AR, besides the empirical precision, the bias caused by carrier-phase multipath should be also fully considered. The coefficients of the ionosphere estimation and GIF model are so large that the multipath errors are accentuated many times. Experiments results indicate that even smoothed or averaged with several hours, the errors of ionosphere estimators could be still over 10 cm; the biases of NL ambiguities with GIF model could be still over 0.5 cycles, especially for BDS GEO satellites.