Undifferenced GLONASS Ambiguity Resolution Over Inhomogeneous Receivers Stations: Introducing Ionosphere Corrections or Resolving Ionosphere-free Ambiguities?

摘要

Undifferenced ambiguity resolution for precise point positioning (PPP) can hardly be implemented for GLONASS data from inhomogeneous stations as GLONASS is based on FDMA (frequency division multiple access) signals. Since receiver hardware biases for all GLONASS satellites at a station are not identical, i.e., inter-frequency biases (IFBs), it is impossible to eliminate them by single differencing between satellites for the FCB (fractional-cycle bias) estimation to enable PPP ambiguity resolution. Although phase IFBs can be modeled with their linear relationship to the channel number, there is no easy approaches to correct for pseudorange IFBs, which thus prevent undifferenced ambiguity resolution across inhomogeneous GLONASS receivers. Geng and Bock (2016) proposed a method where external ionosphere products are introduced to constrain line-of-sight ionosphere parameters in order to reduce the adverse impact of pseudorange IFBs in contaminating undifferenced ambiguity estimates. In contrast, Banville (2016) proposed a method where ionosphere-free ambiguities of a wavelength of about 5 cm can be directly used for ambiguity resolution, as they are not affected by pseudorange IFBs. We can find that either method has its advantages and disadvantages, that is the former needs external ionosphere data which are not readily accessible or available to common users while Global Ionosphere Maps (GIMs) are disqualified due to their poor accuracy over a wide area, and the latter method is riskier when an ambiguity of about 5.3 cm wavelength has to be fixed to integers. Therefore, in this study, we are going to investigate both methods and inspect their performance in undifferenced ambiguity resolution with GLONASS data. We find that GLONASS ionosphere-free ambiguity resolution can achieve a high efficiency if integrated with GPS PPP ambiguity resolution, where about 90% of GLONASS ambiguities in hourly static solutions can be successfully fixed to integers, while 70% otherwise. However, GLONASS ionosphere-free ambiguity resolution in real-time scenarios is impaired by poor satellite orbits and insufficient error mitigations (atmosphere, etc.). On average, 16.8 minutes of observations are required to achieve an ambiguity-fixed solution in real time, even if GPS PPP ambiguity resolution is also carried out. In contrast, combining GLONASS and GPS PPP ambiguity resolution by resolving wide-lane and narrow-lane ambiguities can shorten the convergence time to about 6.2 minutes in real time.