Calibration and analysis of BDS receiver-dependent code biases

Abstract

Different designs of receiver correlators and front ends will cause different biases in terms of the pseudorange. In this paper, we present an algorithm to derive GNSS receiver-dependent code biases by estimating an additional code bias for each satellite based on different orbit and clock products and apply the calibrated biases in user positioning. One-month data from 117 globally distributed multi-GNSS experiment stations are selected to analyze the characteristics of BeiDou Navigation Satellite System (BDS) code biases, including BDS-2 and BDS-3 satellites. It is shown that the BDS code biases on B1I and B3I signals differ by receiver type, receiver model, antenna type, or even unknown factors. The code biases on BDS-2 satellites can reach up to 4 ns from peak to peak, while they are within 2 ns for BDS-3 satellites. The receiver-dependent code biases are calibrated based on the broadcast ephemeris and the precise products. With a correlation coefficient of approximately 0.9, it is demonstrated that the calibrated code biases are in agreement not only among different products, but also with previous result of BDS-2. To validate the effect of the calibrated code biases, we apply them in single point positioning (SPP), precise point positioning (PPP), and real-time kinematic (RTK) positioning. It is proven that with corrected BDS code biases, SPP users will benefit from an overall improvement of 9.5%, 9.9%, and 27.4% at B1I, B3I, and B1I/B3I ionosphere-free frequencies. For the ionosphere-free based positioning using BDS-3 only, the impact of code biases cannot be ignored. The corrected code biases will accelerate the convergence performance in PPP. For RTK applications, especially for single-frequency users, the fixing rate and ratio values will also improve when the code biases are corrected.