Single-Frequency Code-Based RT-PPP GNSS-Aided INS for Connected Vehicles
Eduardo A. O. Borges, Felipe O. Silva, Danilo Alves de Lima, Vinícius M. G. B. Cavalcanti · 2024
Nowadays, precision (and affordable) vehicle navigation is of paramount importance for diverse applications such as Intelligent Transportation Systems (ITS), Connected and Autonomous Vehicles (CAV), Vehicle-to-Vehicle (V2V) communications and Precision Agriculture (PA), requiring advanced techniques like the integration between Global Navigation Satellite Systems (GNSS) and Inertial Navigation Systems (INS), to bridge gaps between periods of satellite signal outages. Furthermore, to optimize the navigation solution and mitigate the so-called Common-Mode Errors (CME), Real-time Precise Point Positioning (RT-PPP) products, such as the ones provided by the International GNSS Service (IGS) in State Space Representation (SSR) format and from the University of La Plata (UNLP) in the form of Regional Ionospheric Maps (RIM), are employed. Based on data from real experimental tests, this study evaluates the performance (in terms of accuracy) of two well-known strategies for fusing INS and GNSS, namely, Tightly-Coupled (TC) and Loosely-Coupled (LC) integrations, in the particular scope of Single-Frequency (SF) code-based RT-PPP, i.e., aiming at low-cost CAV applications in Brazilian territory. We evaluate the degrading impacts associated to (a) different tunings of the measurement uncertainties on the employed estimators; and (b) the loss of signals from the available satellites. As the main contribution of the study, we show that RT-PPP GNSS-aided INS via TC integration is the best performing strategy, delivering horizontal position accuracy better than 2 meters, on average, even when there is only one GNSS satellite available for INS aiding.