ESA GNC Conference Papers Repository
Evaluating differential GNSS techniques for landing the 1st stage of an RLV with a skydiver experiment
As part of the trinational technology demonstration project CALLISTO (Cooperative Action Leading to Launcher Innovation for Stage Toss-back Operations), JAXA, CNES, and DLR are jointly developing key technologies for a future European reusable launch vehicle (RLV). The goal of the project is to build and fly a vertical take-off vertical landing (VTVL) launch vehicle demonstrator that returns to the launch site (RTLS) after its mission. In 2025, there will be ten different test flights with increasing apogee altitudes from several meters to up to 45 kilometers and increasingly complex trajectories from the former Diamant site at Europes spaceport in French Guiana. The GNSS & Navigation Technology group at the German Space Operations Center (GSOC) is responsible for the development, qualification, and operation of the GNSS subsystem that is part of the hybrid navigation system used to estimate the vehicles position, velocity, and orientation. Due to the high demands on the lateral position accuracy, especially during the boost phase immediately prior to landing, a differential code-based GNSS (DGNSS) method is used, which enables horizontal position accuracies in the sub-meter range. The DGNSS system consists of the rover GNSS receiver onboard the vehicle, a reference station located near the landing pad and the uplink of RTCM correction messages from the reference station to the rover GNSS receiver via telecommand. To achieve the lowest possible propellant consumption, the landing boost phase is designed in such a way that the RLV decelerates rapidly until landing and just touches down at almost zero velocity. This approach requires that the altitude of the vehicle above the landing pad is known very accurately in the decimeter range in the final flight phase. It was originally planned to measure the altitude above ground with radar altimeters integrated into the landing legs. However, preliminary studies have shown that radar faces many technical challenges, such as the effects of the engine plume, stress from heat, swirling dirt, and signal reflection by adjacent structures like lightning rod towers. The GNSS & Navigation Technology group is therefore investigating whether the height can alternatively be measured reliably using the differential carrier phase-based real-time kinematic (RTK) positioning method. RTK would provide centimeter-level measurement accuracy of the height once all carrier phase ambiguities have been successfully fixed to integer values. It is advantageous that no additional hardware and infrastructure is required for RTK besides the already existing DGNSS system. However, more effort must be put into the design and placement of the rover antenna on the vehicle, because RTK is more demanding, for example, in terms of exactly knowing the antenna phase center, which in turn requires the careful calibration of the rover antenna. At best, the antenna is already integrated into the structure of the vehicle. RTK is a proven positioning technique and is traditionally used for terrestrial applications such as land surveying, construction or precision agriculture. These applications have in common that they are quasi-stationary and use more or less horizontally leveled geodetic-grade antennas, with the rover and reference antennas at approximately the same height. In contrast, RLV are characterized by high downward velocities and dynamics during the landing boost phase. During the final 2000 meters above the ground, only a few dozen seconds remain until touchdown, during which the ambiguities of a sufficient number of carrier phase measurements must be fixed to integer values in order to calculate a position solution with centimeter-level accuracy. Once an ambiguity has been successfully fixed, the signal has to be continuously tracked until landing. The residual differential tropospheric error due to the rover and reference antennas being at different heights, primarily caused by humidity in the air that cannot be adequately represented by the common troposphere models, may hinder the successful fixing of the carrier phase ambiguities. It is reassuring, however, that this error decreases with decreasing altitude above the landing site, consistent with the accuracy requirement for the height measurement, which is higher the closer the vehicle is to the ground. Other challenges such as the effects of the engine plume passing the rover antenna, vibration from the engine, and multipath have to be overcome to provide accurate and reliable positioning. There is limited experience on how RTK performs in such an application, and to the authors knowledge, there are no reference applications in the literature. Some of the aforementioned challenges are currently being investigated through software simulations and hardware-in-the-loop testing using a GNSS signal simulator. However, to gain an even better understanding of the RTK performance to be expected on an RLV, a skydiver experiment was done in April 2022. The skydivers free flight phase prior to parachute deployment represents the final descent phase of an RLV quite realistically. A special helmet was developed with a multi-frequency GNSS antenna, the attitude of which can be adjusted so that the antenna points upward, depending on the skydivers planned flight pose. The geodetic-grade GNSS receiver, the model used in the CALLISTO project, was stowed in a fanny pack. In addition, a small integrated navigation system consisting of MEMS-based three-axis accelerometers, gyroscopes, and magnetometers, as well as a barometer and another consumer-grade GNSS receiver, was installed on the helmet. It provided reference data for the post-processing analysis. A reference station was operated near the airfield to collect reference measurements. A total of ten flights each from about 4000 meters above ground were done. The skydiver took various poses, such as prone, stand-up, or dive, reaching downward velocities of up to 390 km/h, which is nearly representative of the vertical velocity of the CALLISTO vehicle. Since the differential residual tropospheric error depends primarily on the unmodeled humidity of the air, the effect varies regionally across the globe. Due to the very humid climate in the equatorial regions, the influence is especially large at Europes spaceport in French Guiana and therefore needs to be particularly well understood and analyzed for the application of RTK in the CALLISTO project. The presentation and accompanying paper will describe in detail the application, the experiment, the analyses performed, and the main results.