Kinematic Precise Orbit Determination for real- and near-real-time using Galileo High Accuracy Service – GHASP3 project

Kurzfassung

Over the last decades, Global Navigation Satellite Systems (GNSSs) have proved to be a key technology for onboard Precise Orbit Determination (POD) of space vehicles in Low Earth Orbit (LEO). This has enabled singlereceiver autonomous space navigation, thereby further supporting mega-constellation maintenance, formation flying and collision avoidance manoeuvres. The availability of low-latency/high-accuracy corrections for GNSS precise positioning is fundamental for achieving cm-level accuracy in (quasi) real time. The precise navigation in real time using GNSS broadcast ephemerides has been assessed in Hauschild and Montenbruck (2021), while additional services such as Fugro SpaceStar® could be used for delivering such corrections from Geostationary Earth Orbit (GEO) satellites, albeit with certain limitations. Since January 2023, Galileo High Accuracy Service (HAS) has started providing precise orbit and clock corrections directly via the E6 signal, and this free-of-charge service opens many possibilities for on-board space navigation. In this work, developed under the ESA Technology Development Element (TDE) program, we present the “Galileo High Accuracy for Space Precise Point Positioning” project, i.e. GHASP3 project, collaboration among Delft University of Technology (TU Delft), the German Aerospace Centre (DLR) and the European Space Operations Centre (ESOC). In this research activity, we take the challenge of investigating the purely kinematic POD solution for space-users in LEO when using Galileo HAS corrections, where no orbital dynamical information is used, as well as no geometric constraints are set on the satellite motion. We consider two implementations: a GHASP3-filter and a GHASP3-batch that rely on a Kalman Filter in generalized and batch formulations, respectively. This enables us to study both real-time and near-real-time performances using currently operational missions (e.g. Sentinel-6A, Sentinel-3B, and GRACE-FO), thus making use of real world GNSS code and phase data acquired on board. When looking at GNSS-based kinematic POD, additional challenges arise, where both ‘quality’ and ‘latency’ of corrections become fundamental for Precise Point Positioning (PPP)-like performance. The same holds for the data pre-processing and quality control, based here on the Detection, Identification, Adaptation (DIA) theory developed at TU Delft. For the activity, we utilized HAS corrections – acquired via NTRIP for the entire month of July 2023 – in order to assess the kinematic solutions. We first present kinematic results for some ground-based users (e.g. DLF1/REDU stations), and then we compute precise orbits for space-based users. Since HAS does not provide satellite phase biases yet, we later relax this assumption by utilizing a few alternative sources of real-time corrections or adopting IGS precise products. In this way, we can investigate numerically ambiguity-fixed POD performance based on Least-squares AMBiguity Decorrelation Adjustment (LAMBDA) method and its recent v4.0 implementation. Overall, the use of an uncombined and undifferenced formulation in GHASP3 software for PPP/POD provides a large flexibility for assessing multi-GNSS/multi-frequency solutions, while it also brings many benefits for scientific applications. These applications are briefly discussed, for instance in relation to the estimation of iono-plasmasphere electron content or to support the real-time monitoring of time offsets in multi-GNSS constellations.