Method and assessment of BDS triple-frequency ambiguity resolution for long-baseline network RTK

Abstract Fast and reliable ambiguity resolution (AR) over long baselines is one of the major challenges for large-scale GNSS network real-time kinematic (RTK) positioning. In the dual-frequency case, wide-lane (WL) ambiguities are usually fixed firstly, followed by the procedure of determination of the narrow-lane (NL) ambiguity integers based on the ionosphere-free model. For long-baseline network RTK, there are still some problems faced, for instance the WL AR still needs to take over tens of epochs due to the influence of pseudorange noises and multipath effects; the NL AR model is weak since more parameters are introduced, thus much time is needed to guarantee the precision of float ambiguity solutions. Along with the deployment of modernized GPS, BDS and other GNSS, the triple-frequency AR with real data gradually becomes feasible. Many literatures have proven the third signal will make AR faster and more reliable. In this paper based on the characteristics of network RTK, a modified ionosphere-free model was proposed to improve EWL/WL AR performance for long baselines, followed by the comparison with several commonly-used EWL/WL AR models. The NL AR performance with triple-frequency and dual-frequency observations was also tested. In NL AR, a partial AR strategy is adopted to weaken the negative influence of new-rising or low-elevation satellites. Experiments were conducted using real BDS triple-frequency data on 74 km and 241 km baselines. Results show that the modified ionosphere-free model can balance the ambiguity precision and ambiguity bias better for the EWL/WL AR. The NL AR with triple-frequency observations also performs significantly better than that in dual-frequency mode, with 29.9% and 32.9% improvement in time-to-first-fix (TTFF) for the 74 km and 241 km baselines respectively.