Real Time GPS Positioning of LEO Satellites Mitigating Pseudorange Multipath Through Neural Networks

摘要

A method for real time positioning of LEO satellites using dual frequency GPS receivers is presented. It is based on an apriori on ground estimation of an irregular pseudorange multipath map computed with the help of a Self-Organizing Map neural network algorithm. The generated Self-Organizing Multipath Map (SOMM) characterizes the multipath environment of the satellite. The accuracy obtained with this approach improves the one that can be achieved with a regular grid map not adapted to the multipath variations. This apriori estimation allows a real time correction of the code observables with a number of parameters which is affordable for space vehicle applications in terms of CPU and memory usage. The LEO satellite would correct the observable directly with the SOMM information in real time. Afterwards, the position is obtained from the corrected pseudoranges by means of standard least mean squares. Precise IGS clocks and orbits have been used to measure the impact of these corrections in the positioning. The final navigation solution reduces the RMS of the error by about 40%-50% for SAC-C satellite (obtaining errors of 90 cm.) and by about 25%-35% for CHAMP satellite (obtaining errors of 70 cm.).