Infoscience

Student project

GPS/INS Integrity in Airborne Mapping

The quality of the laser point cloud georeferencing in airborne laser scanning missions is largely related to the quality of the GPS solution. The latter is obtained by post- processing the differential carrier-phase measurements in order to reach the required accuracy. This implies that errors or unacceptable quality in the gathered data that cause problems for the ambiguity resolution in the post- processing step are detected much later. The objective of this thesis is to investigate new concepts of GPS data quality monitoring and to improve the GPS solution by using RAIM and WAAS/EGNOS integrity enhancement techniques. To do that, quality check algorithms based on indicators such as the signal-to-noise ratio, the cycle slip detection results or the phase tracking loop output are proposed and successfully tested. Furthermore, a new global quality check algorithm based on RAIM and cycle slip detection has been designed and tested with a focus on the chances to resolve correctly the ambiguities during the carrier-phase post-processing. The algorithms are implemented in a real- time quality check tool developed in a C/C++ environment whose performance shows that the provided quality indications enhance the GPS integrity by providing crucial information on the signal quality during the flight. This information enables problematic epoch identification and warns immediately the mission operator about problematic flightlines that should be flown again. This avoids final product quality degradation or expensive mission repetition. The thesis also presents the design of an RTK- GPS on-board solution via radio communication channel. The design has been tested during a flight and the results show that a sub-decimetric accuracy can be reached by this mean. The potential of using such a solution is high since this provides ultimate integrity test for phase data. Also, as the final laser point cloud has been georeferenced quite accurately using the real-time sensor observations and Kalman filtering, the economical gain of avoiding post- processing is substantial.

Related material