With the renewed interest in lunar surface exploration, the European Space Agency envisions to stimulate the creation of lunar communications and navigation services (LCNS) to enable, among others, autonomous navigation capabilities for lunar rovers. As the number of satellites foreseen in such a service is much smaller compared to Earth-based global navigation satellite systems, different complementary technologies are pursued to improve the attainable navigation accuracy for lunar rovers. One way to improve the position accuracy provided by the LCNS satellites is to constrain their vertical position using a high resolution digital elevation model (DEM). This article presents the results of a variance covariance analysis of an extended Kalman filter implementation in which the LCNS ranging measurements are used together with the altitude provided by a DEM from the Lunar Orbiter Laser Altimeter instrument of the Lunar Reconnaissance Orbiter. Assuming a realistic orbit determination and time synchronization accuracy of the LCNS satellites, the usage of a navigation-grade inertial measurement unit and an oven-controlled crystal oscillator, a 3-sigma position accuracy of less than 10 m can be obtained. Furthermore, the availability is substantially improved as the DEM-aided solution enables a position solution in case of only 3 visible satellites.

In GNSS-based navigation onboard Low Earth Orbit (LEO) satellites, typical accuracy requirements are 10 cm and 0.1 mm/s for 3D position and velocity, respectively. Previous works have shown that such performance is achieved by including Galileo measurements in the estimation process. Here, we aim to evaluate the impact of employing State Space Representation (SSR) corrections, i.e., GNSS satellite orbit, clock, and biases, to be applied to the broadcast ephemerides. In this framework, the Precise Onboard Orbit Determination (P2OD) software (SW) tool developed at the University of Padua (UNIPD) is used to investigate the needs of onboard navigation. The UNIPD SW employs an Extended Kalman Filter (EKF) using a reduced-dynamics approach. The force model implemented is adapted to onboard processing, and empirical accelerations are included to take into account residual force mismodeling. Actual observation data from the LEO Sentinel-6A satellite are processed along with SSR corrections from the CNES service. Galileo-based solutions are compared to ground-based POD reference orbits. The analysis suggests that the use of SSR corrections provides sub-decimeter and below 0.1 mm/s accuracies in 3D position and velocity, respectively. Such results indicate a P2OD solution quality close to that achievable by adopting precise orbits and clocks.

With the renewed interest for lunar surface exploration, the European Space Agency envisions to stimulate the creation of lunar communications and navigation services (LCNS) to enable, among others, autonomous navigation capabilities for lunar rovers. As the number of satellites foreseen in such a service is much smaller compared to Earth based global navigation satellite systems (GNSS), different complementary technologies are pursued to improve the attainable navigation accuracy for lunar rovers. One way to improve the position accuracy provided by the LCNS satellites is to constrain the vertical position using a high resolution digital elevation model (DEM). This article presents the results of a variance covariance analysis of an extended Kalman filter (EKF) implementation in which the LCNS ranging measurements are used together with DEM from the LRO LOLA instrument. Assuming a realistic orbit determination and time synchronization (ODTS) accuracy of the LCNS satellites, the usage of a navigation grade IMU and an oven controlled crystal oscillator (OCXO), a 3-sigma position accuracy of less than 10 meters can be obtained. Furthermore, the availability is substantially improved as the DEM aided solution enables a position solution in case of only 3 visible satellites.