Using Radar to Identify Drones

Lead Research Organisation: University of Birmingham
Department Name: Electronic, Electrical and Computer Eng

Abstract

The project will aim to use staring radar to distinguish between birds and drones in airspace. Deep learning algorithms will be used to enable the radar system to detect, track and classify individual birds and drones. This will primarily involve working with experts from within EESE (Prof Chris Baker and Dr Mike Antoniou), with a possible informal collaboration with colleagues in Biosciences. The interdisciplinary nature of this project will allow two areas of research to be brought together in order to contribute novel methods and data to the community. This will benefit the scientific community (understanding of how birds' flight characteristics vary with changing environments, and a greater understanding on how information is encoded in radar backscatter signals) and in industry (applications to the monitoring of airspace, vehicular radar, vitalsign health monitoring, military and security surveillance etc).

Provisionally, the primary objectives of the research project will be as follows:
To develop a synthetic simulation environment in which bird and drone data can be input, allowing a deep learning algorithm to be created and tested on a variety of data, to parameterise the model and assess its performance. To develop an automatic discrimination method which will enable the staring radar to identify whether a visible target is a bird or a drone, and to distinguish between different breeds of bird.
To evaluate data to determine whether it is possible to detect individual birds from within a flock, and to find the smallest size of bird that may be detected.
During the course of this project, I hope to develop a deep-learning algorithm which can be used with staring radar in order to classify the objects detected in airspace.

The key outcomes of this research will be developing an instrument to differentiate between birds and drones, and between d individual birds. Colleagues in Biosciences will then be able to use this research to estimate total bird flying
biomass for a city, and analyse the effects of urban development on the flight and migration activity of birds.

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/N509590/1 01/10/2016 30/09/2021
2109392 Studentship EP/N509590/1 01/10/2018 30/09/2021 Holly Anita Dale