EU remote
Master's thesis: AI Driven 3D Radar for Drone Detection with 6G ISAC
About this role
We are looking for a dedicated master’s student to join us at RISE, in cooperation with the Institute of Microwave and Photonic Engineering (IHF) at TU Graz. The thesis is part of RADAR PAN-SWEDEN, which aims to validate distributed radar sensing for drone detection at test sites across Sweden. The work combines radar signal processing, edge AI and 5G connectivity, and is carried out in close collaboration with researchers at TU Graz.
Background and Purpose Small commercial drones increasingly pose risks to airports, critical infrastructure and public events. They are small, fly low and move slowly, which makes them hard to detect with conventional radar systems designed for larger targets. 6G Integrated Sensing and Communication (ISAC) opens new possibilities. Future mobile networks are expected not only to transmit data but also to sense their environment, so that radar nodes distributed across the network can observe the same airspace from several angles.
Combined with AI at the network edge, such systems could detect and classify drones in real time, close to where the data is generated. However, moving from simulation to real-world operation remains a challenge. Real radar data, the constraints of embedded hardware, synchronization between nodes and reliable data transfer over cellular networks all affect how such a system performs in practice. Thesis Description The goal of this thesis is to take an existing end-to-end radar emulation framework into the real world, using physical 77 GHz FMCW radar modules.
You will start with a review of 6G ISAC principles, distributed multistatic radar architectures and RF-based drone detection, and then specify and integrate the radar hardware together with researchers at TU Graz. You will plan and carry out measurement campaigns to record raw radar data and micro-Doppler signatures of commercial multi-rotor drones and background clutter. Using this data, you will implement and benchmark the radar processing chain, including range, Doppler and azimuth FFTs and adaptive CFAR detection, on embedded edge hardware such as NVIDIA Jetson.