PhD Studentship: Low Cost Acoustic Camera Designs and Algorithms for Tracking Swarms of Drones / UAVs

Tracking and detection methods for UAVs and drones typically involve human in the loop, with technology including long range radar, short range radar, cameras, infra-red camera, RF communication technology and acoustics. 

At shorter range, cameras and acoustic microphones can detect the drones and provide tracking data to a wide distributed sensor network, which includes management and path prediction. This allows time for a person to intercept and address the drone behaviour. Cameras need to know which way to point and in cluttered urban environments suffer performance degradation.

In this PhD project, the ambition is to study the performance of acoustic camera designs for tracking narrow band noise sources, using arrays of cheaper microphones with lower signal to noise ratios than expensive options. The intention is to simulate a wide sensor network of tracking devices. Beamforming algorithms will be optimised for single drones, multi-drones and swarm options, looking for differences in the mass that could indicate variants.  

The department has a large UAV laboratory, with recent experience of target tracking algorithm development in this field. We wish to further this research by looking specifically at the sensors and sensor management in terms of technical capability, signal to noise problems and blocking / reflections or even contamination of the signal so that these can be included in the simulation environment. 

It is anticipated that the project will involve both simulation development, largely in Matlab / Simulink and experimental measurements of real UAV drones with microphone arrays, looking into directional measurements. Hence an interest in UAV movements would be a significant advantage. 

The student will be expected to set up a real-time Simulink model of the drone noise with sensors distributed over a 2D area. Beamforming and focusing methods will be used on simulated microphones to develop a minimum number and distribution or whether distant microphones can be compared using a GPS timebase. 

Research questions to answer include the beamforming for low quality signals, the implications of this on the wider statistical prediction methods, the ability to categorise the drone by type using either the spectral characteristics or machine learning and how this provides a statistical model of the sensor.  

In addition to the sound acquisition, machine learning may be critical to identify drone types from a training database but we might also look at radio frequency signals or radiation of frequency data including electromagnetic signatures and communication protocols as additional sensors using inexpensive software defined radio platforms.

Supervisors

Primary supervisor: Dan O'Boy

Secondary supervisor: Matthew Coombes

Entry requirements

Students should have, or expect to achieve a 2:1 undergraduate degree in a relevant subject.

An interest in either acoustics, vibration, UAV, flight control and tracking, radar, beamforming or signal processing would be an advantage.

How to apply

All applications should be made online via the above ‘Apply’ button. Under programme name, select AACME / AAE Department of Automotive and Aeronautical Engineering. Please quote the advertised reference number: * AACME-24-033 * in your application.

Funding Details

Funding Comment

The 3 year studentship provides a tax-free stipend of £19,237 per annum, plus tuition fees.

Advert information