Acta Geodaetica et Cartographica Sinica ›› 2026, Vol. 55 ›› Issue (6): 1047-1057.doi: 10.11947/j.AGCS.2026.20250404

• Geodesy and Navigation • Previous Articles    

Unified estimation and correction method for multi-form code bias using BDS-3 data

Jingzhu ZHAO1,2(), Chuang SHI2,3, Lei FAN2,3(), Shiwei GUO2,3, Tao ZHANG1,2   

  1. 1.School of Electronic and Information Engineering, Beihang University, Beijing 100191, China
    2.Key Laboratory of Navigation and Communication Fusion Technology, Ministry of Industry and Information Technology, Beijing 100191, China
    3.School of Space and Earth Sciences, Beihang University, Beijing 102206, China
  • Received:2025-09-28 Revised:2026-06-01 Published:2026-07-28
  • Contact: Lei FAN E-mail:jingzhu@buaa.edu.cn;bhflei@buaa.edu.cn
  • About author:ZHAO Jingzhu (2000—), female, PhD candidate, majors in GNSS precise data processing. E-mail: jingzhu@buaa.edu.cn
  • Supported by:
    The National Natural Science Foundation of China(42274041; 42404038; 41931075);Young Elite Scientists Sponsorship Program by CAST(YESS20230249)

Abstract:

Global navigation satellite system (GNSS) code biases are primarily characterized by three forms: observation-specific signal bias (OSB), differential code bias (DCB), and inter-frequency clock bias (IFCB). Precise processing of these code biases is critical for achieving consistency in observation data across different GNSS frequencies. However, due to differences in estimation models, error handling strategies, and reference definition, code bias products from different sources and in different forms exhibit inconsistencies. To address this issue, this study derives a compatible and unified functional model for estimating different forms of code biases and establishes a unified correction method for these code bias forms. Using global observation data from the BeiDou-3 Global System (BDS-3), the three forms of code bias (OSB, DCB and IFCB) are estimated and the results are evaluated. Results demonstrate that systematic biases exist between the code biases estimated from the uncombined model and external products, which are attributed to inconsistencies in the clock datum. Additionally, high consistency is observed among the OSB, DCB, and IFCB estimates, with root mean square (RMS) differences all within 3.0×10-3 ns. The estimated code biases are further applied to precise point positioning (PPP), achieving unified correction of multi-form code biases at the clock offset level. Experimental results demonstrate that, compared with using external OSB products, the code biases estimated in this study led to average reductions in the RMS of 3D positioning errors during the convergence period by 4.9% and 9.3% for the B1I/B2a and B1C/B2a frequency combinations, respectively. All three code bias forms demonstrated consistent positioning performance during both convergence period and after convergence. Therefore, in practical applications, users can flexibly select any form of code bias according to their data conditions to achieve consistent and high-precision positioning results.

Key words: observation-specific signal bias, differential code bias, inter-frequency clock bias, undifferenced and uncombined, precision point positioning, BDS-3

CLC Number: