Characterization of GNSS Line Bias and BDS-3 Attitude Determination With Single-Differenced Observations From Common-Clock Receiver

Global navigation satellite system (GNSS)-based attitude determination with joint use of the single-differenced (SD) model and common-clock receiver has attracted much attention recently. In this contribution, for the first time, we investigate the impact of antenna type and coaxial cables on the GNSS line bias (LB). The experimental results demonstrated that no apparent impact of antenna type is observed on the stability of the LB. The phase LB could, however, vary significantly if large difference exists between the lengths of the involved two cables of the same type and specification. Moreover, we found that large variations in the phase LB may be present during a continuous long-term observation period even if two cables with the same length are used. Then, we evaluate the performance of BeiDou global navigation satellite system (BDS-3) attitude determination with SD and double-differenced (DD) models, using both static and car-driven kinematic data collected with a Trimble BD992 common-clock receiver. Particularly, for the first time, the performance of the new B1C/B2a signals is presented and compared with B1I/B3I signals. Experimental results demonstrated that compared with the DD model, the SD model provides improved ambiguity resolution (AR) success rate and attitude accuracy. The achievable AR success rate and yaw accuracy are slightly improved, whereas the pitch accuracies are remarkably improved by approximately 37.9%–50.0% and 46.0%–53.1% for the two datasets. Moreover, the results of static data have revealed that the BDS-3 B1C solution delivers comparable attitude accuracy with B1I solution, whereas the B1C/B2a solution provides lower attitude accuracy than B1I/B3I solution.