Millimetre wave radar signatures of drones

The detection of small unmanned aerial vehicles (UAVs), or drones, is a growing requirement both for the management of intentional drone activity and for the detection of unwanted drones which might cause a risk to security, safety or privacy. Radar is one of the most promising sensors with which drones can be detected and classified by using distinct radar signatures such as radar cross section (RCS) and micro-Doppler (produced by the rotation of the propeller blades). Research in the Millimetre Wave Group has shown that millimetre wave radar may offer some advantages in drone detection, tracking and classification in terms high Doppler sensitivity and compact system size. This PhD project aims to deepen the understanding of the millimetre wave radar signatures of drones and develop techniques which could be applied in a drone detection system. The scope of research topics includes:-
1. Develop statistical fluctuation models to analyse RCS characteristics of drones at millimetre-wave frequencies using data available at St Andrews.
2. Characterise RCS values of drones in terms of fuselage and propeller material, size, shape and aspect angles.
3. Study polarisation dependencies of the RCS values of the drones at millimetre-wave frequencies by using the available data and acquiring more data by using radars available at St Andrews.
4. Develop EM scattering models of commercial drones to study and characterise the micro-Doppler signatures produced by the propellers. Understanding of the features will be further strengthened by lab-based controlled environment experimental data collection and analysis.
5. Wavelet analysis of the micro-Doppler signatures of drones. Develop wavelet signature based drone classification algorithm for real-time application.
6. Inverse Synthetic Aperture Radar (ISAR) imaging of drones using high resolution range profiles (HRRP) data. Study the feasibility of distinguishing attached payloads and different models of drones using ISAR data.
7. Develop techniques for Neural Network based drone classification algorithm for real-time application. Extensive data is already available at St Andrews to create a training database. The candidate will also design and set up experimental trials to make the database larger and more diverse in terms of number of drones and surrounding environment.