The project was about how to use a distributed, low-profile ground station observation network (GENSO) to determine the orbit of low-earth orbiting (LEO) satellites. Below you can find some documents to download, and the thesis summary.
For the people interested, you can have a look at my thesis (pdf 8.7 MB).
In order to communicate with these satellites, the orbit of these satellites has to be known. Currently, data about the orbits is available, but the availability of this data is not guaranteed for the future. With GENSO being a world-wide satellite observation network, it is a perfect candidate to generate this orbit data, by performing orbit determination itself. How satellite orbit determination can be performed with GENSO, is the research topic of this thesis.
Because GENSO was being built during this research by the GENSO developers, but not operational yet, a simulation model of GENSO was developed. With this simulation model, the orbit determination methods that were developed were tested. Two different strategies were developed to perform orbit determination during the entire lifetime of the satellite, starting from its launch. Despite the relatively low-end communication resources of GENSO, the simulations show that GENSO is able to perform orbit determination by itself, and can be selfsupporting in obtaining data about the orbits.
The first strategy, called static mode orbit determination, covers the situation where no prior information about the orbit is available. The orbit data is reconstructed out of the observations made by the GENSO observation network. After 2 days of running the orbit determination process, usable orbit data is generated.
The second strategy, called tracking mode orbit determination, covers the situation where initial orbit data is available. A smart search algorithm tests different orbits, to see which orbit fits best to the orbit of the satellite. Usable orbit data is generated after 5 days. The search algorithm can be categorized as a hill-climbing trial-and error search algorithm, based on the ideas of reinforcement learning.
In this research is shown that it is possible to perform orbit determination for educational satellites with GENSO, by means of experiments performed with a simulation model of GENSO. It is shown that the orbital parameters determined by GENSO are not as precise as the ones that are currently provided, but the accuracy is sufficient to make communication with the satellites possible.