In order to apply for any of the following theses or internships, the candidate must have no more than 3 exams left.
At present, the orbit determination of deep space missions relies mainly on Earth-based radiometric measurements, namely ranging, Doppler, and DDOR. These are derived from the properties of the radio link between the spacecraft and one or more ground stations on the Earth. The main sources of noise affecting the radio link are: interplanetary plasma, Earth’s troposphere and ionosphere, thermal noise in the electronics.
The objective of this project is to develop a Python tool to quickly evaluate the quality of the radiometric measurements acquired at the ground stations, without the need of a detailed orbit determination analysis. The candidate will have to retrieve and load all the relevant inputs, including: radiometric measurements, meteorological data, station configuration, spacecraft telemetry. Then, the most important parameters affecting the link quality will be computed and displayed. An automatic test report will be generated.
Synthetic images are crucial in training AI-based image processing techniques since the availability of real image examples for deep space missions is limited. This thesis explores using Generative Adversarial Networks (GANs) to bridge the domain gap between synthetically generated images created with 3D computer graphic software and real mission imagery. The synthetic images generated will be used as a base for the GAN algorithm to add the noise extracted from real images. The improved synthetic images generated through this approach will subsequently be used to train an existing image detection algorithm based on Convolutional Neural Networks (CNN) to detect particles orbiting small bodies.