GNSS can provide high accuracy via real-time-kinematic (RTK) positioning, but RTK currently works only if there is a free line-of-sight between user and GNSS satellite. In urban areas the GNSS signal is often obstructed or reflected. The availability of a RTK solution in urban areas is less than 50-70 % (depending on the type of the buildings), which renders RTK virtually useless for safety critical applications. For example, we can currently not imagine that an UAV flies autonomously in a city, e.g. for traffic monitoring.
This project targets to increase the RTK availability in urban areas by using GPS+Galileo L1/E1 and L5/E5a signals and by inertial aiding. A cost efficient microelectromechanical (MEMS) gyro and accelerometer (a so-called inertial measurement unit – IMU) is used to detect cycle-slips within the GNSS carrier phase measurements. Cycle-slips are the main reason for the low RTK availability in urban areas. A GPS/Galileo RTK module is developed which is integrated with an ultra-tightly coupled GNSS/IMU receiver, to achieve continuous GNSS signal tracking even if obstructions are present. To mitigate multipath the principle of a synthetic antenna aperture is used. The Kalman filter is extended based on IMU calibration campaigns and a parameter sensitivity analysis. This shall allow to use more cost efficient IMUs or to bridge longer GNSS signal outages (e.g. in an underpass).
For verification, a ring-laser IMU based reference system is integrated within a measurement vehicle. This allows establishing the ground truth with ~ 1 cm accuracy. The measurement vehicle also carries competitive solutions used for comparisons. A fish-eye camera allows identification of GNSS signal degradations and is used to optimize the developed algorithms.