High-Rate Attitude Determination of Moving Vehicles With GNSS: GPS, BDS, GLONASS, and Galileo

We investigate the long-term performance and improvement of high-rate multiconstellation Global Navigation Satellite System (GNSS) or multi-GNSS attitude determination in a dynamical environment and evaluate how much each GNSS contributes to such an improvement with applications to land vehicle navigation. We build a multiantenna GNSS system with three nondedicated receivers and conduct two experiments with quad-constellations in GNSS-friendly and GNSS-challenged environments on January 19 and 23, 2018, respectively. A high-grade inertial measurement unit (IMU) provides a reference of comparison for dynamic attitude solutions. In the static case, the global positioning system (GPS), BeiDou navigation satellite system (BDS), and GLObal NAvigation Satellite System (GLONASS) single-constellation solutions have basically the same performance for both experiments. In the dynamic case, GPS achieves the best accuracy in the experiment on January 19, while on January 23, GPS and GLONASS achieve very similar accuracy and significantly outperform BDS in the pitch and roll components. The different dynamic performances of these single constellations are primarily attributed to their differences in observation noise level and satellite geometry strength. By combining GPS, BDS, GLONASS, and Galileo, the multi-GNSS solution performs significantly better than any single-constellation solution. The average improvements in the dynamic cases are 26.2%, 40.7%, and 36.0% on January 19 and 19.2%, 30.0%, and 36.2% on January 23 in the yaw, pitch, and roll components, respectively. The attitude solutions on January 19 are always better determined than on January 23 in the dynamic cases, which should clearly indicate the importance of combining multiple GNSS constellations and the benefit of more observed satellites and a stronger satellite geometry in GNSS-challenged environments. We also introduce the concept of "influence rate" and quantitatively analyze the contribution of each GNSS system to the combined attitude solution.