Rapid ambiguity resolution over medium-to-long baselines based on GPS/BDS multi-frequency observables

Medium-long baseline RTK positioning generally needs a long initial time to find an accurate position due to non-negligible atmospheric delay residual. In order to shorten the initial or re-convergence time, a rapid phase ambiguity resolution method is employed based on GPS/BDS multi-frequency observables in this paper. This method is realized by two steps. First, double-differenced un-combined observables (i.e., L1/L2 and B1/B2/B3 observables) are used to obtain a float solution with atmospheric delay estimated as random walk parameter by using Kalman filter. This model enables an easy and consistent implementation for different systems and different frequency observables and can readily be extended to use more satellite navigation systems (e.g., Galileo, QZSS). Additional prior constraints for atmospheric information can be quickly added as well, because atmospheric delay is parameterized. Second, in order to fix ambiguity rapidly and reliably, ambiguities are divided into three types (extra-wide-lane (EWL), wide-lane (WL) and narrow-lane (NL)) according to their wavelengths and are to be fixed sequentially by using the LAMBDA method. Several baselines ranging from 61 km to 232 km collected by Trimble and Panda receivers are used to validate the method. The results illustrate that it only takes approximately 1, 2 and 6 epochs (30 s intervals) to fix EWL, WL and NL ambiguities, respectively. More epochs' observables are needed to fix WL and NL ambiguity around local time 14:00 than other time mainly due to more active ionosphere activity. As for the re-convergence time, the simulated results show that 90% of epochs can be fixed within 2 epochs by using prior atmospheric delay information obtained from previously 5 min. Finally, as for positioning accuracy, meter, decimeter and centimeter level positioning results are obtained according to different ambiguity resolution performances, i.e., EWL, WL and NL fixed solutions. (C) 2016 COSPAR. Published by Elsevier Ltd. All rights reserved.