Galileo represents the European contribution to the Global Navigation Satellite System (GNSS) family. It delivers highly accurate and guaranteed global positioning and timing services at worldwide level. Galileo provides two major differentiators with respect to other GNSS systems: High Accuracy Service (HAS) and Open Service Navigation Message Authentication (OSNMA). This study focuses on Galileo HAS for which Initial Service was declared operational on January 24, 2023. Galileo HAS provides free access to precise orbit, clock and bias corrections to enable Precise Point Positioning (PPP). This study gives a systematic evaluation of Galileo HAS performance during its first three years of operations is presented as observed by the Galileo Reference Centre (GRC). GRC provides EU Agency for Space Programme (EUSPA), and other EU stakeholders, with an independent system to evaluate the quality of the GNSS services.

Galileo HAS corrections are disseminated through both satellite and terrestrial channels. Galileo users may retrieve HAS data directly via the E6-B Signal in Space (SIS) component [HAS SIS] or through the Internet using NTRIP protocol [HAS IDD]. Performance assessment is conducted in accordance with the official service definition document. Metrics include correction accuracy, service availability, and user-level positioning performance. GRC reference products are employed to validate HAS orbit, clock and bias corrections. In addition, results from the Galileo HAS User Terminal (HAUT) hosting the Galileo HAS Performance Characterization User Algorithm are analyzed. Convergence times for PPP solutions are also investigated.

The study demonstrates that HAS orbit and clock corrections exhibit root-mean-square errors at the centimeter level, while code bias corrections demonstrate long-term stability. Service availability remains consistently high across the service modes. Galileo HAS routinely delivers decimeter-level PPP positioning with 15-25 cm horizontal and 20-30 cm vertical accuracy at the 95% percentile under open sky conditions within the service area. Convergence time in multi‑constellation static scenarios remains variable, with performance enhancements expected during the HAS Phase 2.

The three-year assessment provides a cumulative, multi-year, quantitative evidence of the robustness and reliability of Galileo HAS. The service consistently delivers accurate corrections and enables PPP solutions with decimeter level positioning accuracy. These findings confirm the Galileo HAS capability to support applications requiring high-precision positioning.

Various applications including smart and connected mobility services and precise agriculture are highly dependent on reliable and continuous PVT (Position-Velocity-Time) information. However, to achieve centimeter-lever accuracy in real-time PPP (Precise Point Positioning), the positioning engine necessitates streams of precise orbit and clock products.
Recently, the release of Galileo High Accuracy Service (HAS) raised high interest in real-time, stand-alone precise positioning. Notwithstanding HAS products are reliable and publicly available, still exhaustive programming it is required to extract them successfully, formulate and feed appropriately the positioning engine. Moreover, assuming HAS corrections have been successfully implemented, there are situations that PVT quality can deteriorate rapidly, particularly in urban and deep-urban canyons, where obstructions and multipath affect severely satellite signal reception.
In this context, this study aims to address the problem via an integrated, tightly coupled Kalman filter approach, featuring enhanced PPP GNSS data (multi-frequency / dual-constellation signals and HAS corrections) and augmentations from inertial observables. In particular, considering the high importance of PVT continuity in safety critical mobility services, special attention is paid in computing and monitoring PVT continuity metric.
The study provides a detailed presentation of HAS orbit and clock products and their format types. Implementation of the HAS corrections resides on GINAV open-source software due to its core PPP structure and scalability features. The paper discusses the software development for parsing the HAS corrections to serve GINAV and the issues encountered. The PVT trajectory results obtained from the kinematic analyses of a number of dedicated kinematic tests are presented and evaluated in terms of position accuracy, convergence and availability. Finally, the study investigates the contribution of HAS corrections in PPP GNSS / INS integration to overcome (to an extent) position discontinuities in harsh kinematic environments. Both solutions are compared against a tactical grade GNSS/INS reference solution.

The Galileo High Accuracy Service (HAS) provides free Precise Point Positioning (PPP) corrections for Galileo and GPS satellites, via the Galileo E6-B signal and the internet. The service was launched in January 2023 in the IS (Initial Service) phase and has significant potential for use, being independent of any terrestrial infrastructure. The recent emergence of commercial receivers compatible with HAS necessitates testing the service at the SL1 level on such receivers to determine the trust domains and usage limits of the positioning obtained. In this study, an analysis of the accuracy of HAS corrections in the Moldova region of Romania is proposed. The tests were conducted in static and kinematic modes, using the SLAGEN S9 Serial GNSS Receiver, during the period from October to December 2025. The objectives were to determine the convergence period for dual-constellation, GPS-only, and Galileo-only scenarios, as well as to evaluate the solution's accuracy. The tests were conducted under different satellite geometry conditions (DOP). The results obtained for the Iași area, Moldova, Romania, indicate a horizontal and vertical accuracy of 95% at 20 cm and 31 cm, respectively, for static positioning in the dual-constellation scenario. The study also evaluates the convergence times required to achieve target positioning performance in dual-constellation, GPS-only, and Galileo-only scenarios.