Tightly coupled integration of multi-GNSS PPP and MEMS inertial measurement unit data

Abstract

Precise point positioning (PPP) using the Global Positioning System (GPS) is widely recognized as an efficient approach for providing precise positioning services. However, its accuracy and reliability could be significantly degraded by unexpected observation discontinuities and unfavorable tracking geometry which are unavoidable, especially in severe environments such as city canyons. Therefore, in the last decades inertial navigation system (INS) has been integrated to overcome such drawbacks. Recently, multi-Global Navigation Satellite Systems (GNSS) were applied to enhance the PPP performance by appropriate usage of the increased number of satellites. We present a new approach to tightly integrate the multi-GNSS PPP and INS together in the observation level. The inter-system bias and inter-frequency bias of multi-GNSS and the hardware errors of INS sensors are estimated to improve the position accuracy and to shorten the convergence time of PPP. In order to demonstrate the impact of multi-GNSS observations and INS data on the derived position, velocity, attitude, and the convergence time of PPP, the new approach is validated through an experimental test with a set of land vehicle data. The results show that the position accuracy can be improved by multi-GNSS and INS significantly, but very little improvement in velocity and attitude is achieved. The position root-mean-square improves from 23.3, 19.8, and 14.9 cm of the GPS PPP/INS tightly coupled integration (TCI) solution to 7.9, 3.3, and 5.1 cm of multi-GNSS PPP/INS TCI in north, east, and up components, respectively. Furthermore, GNSS outages are simulated and their effect on the performance of multi-GNSS PPP/INS TCI is investigated to demonstrate the contribution of the multi-GNSS PPP/INS TCI during GNSS outages. In addition, the convergence test also shows that both multi-GNSS and INS can improve the PPP convergence performance noticeably.