Munich Aerospace PhD Scholarship on Machine-Learning Based Dynamic Radio Resource Allocation for Future Satellite Systems

Munich Aerospace PhD Scholarship on Machine-Learning Based Dynamic Radio Resource Allocation for Future Satellite Systems

The Chair of Signal Processing from the Bundeswehr University Munich is one of the few academic research institutes in Germany with a focus on satellite communications. Its expertise includes several topics related to radio resource management (spectral-efficient techniques, multiple access for IoT, spectrum management) and the integration of satellites into 5G networks. Its international team offers an ideal research environment for a successful and timely completion of the Munich Aerospace PhD programme.

The research works will be conducted in the group “Machine Learning for Network Management and Resource Allocation in Future Satellite Systems” led by Prof. Andreas Knopp. In cooperation with the DLR Institute of Communications and Navigation, challenges related to the design of 5G and Beyond satellite systems should be addressed. The research work focuses on the management of complex non-terrestrial networks involving flexible digital satellite payloads in various orbits (e.g. LEO, GEO). Due to the high-dimensionality of the related optimization problems, the supervision of the resources at the network and the physical layer level requires new approaches. In this context, machine learning techniques will be investigated to enable the deployment of a self-organizing space system.

The PhD student will explore novel approaches for the dynamic allocation of radio resources in next-generation satellite systems. Deep reinforcement learning (DRL) will in particular be considered for the automated optimization of the available power, bandwidth and time resources given the traffic demands. In light of the multilayered architecture under study, distributed solutions implying the cooperation of several agents in the decision-making process will especially be investigated. The theoretical performance limits of the proposed approaches will be analyzed. Moreover, their complexity will be assessed to guarantee the feasibility of their deployment in practical systems.

Interested? Please send us your application including relevant documents (cover letter, CV, diplomas, transcript of records) in PDF format to office.sp@unibw.de. The application deadline is May 27, 2021.

You can download all the details about the Munich Aerospace PhD scholarship here.