Localisation of Drone Controllers from RF Signals using a Deep Learning Approach

Shorten, David and Srivastava, Saket and Murray, John (2018) Localisation of Drone Controllers from RF Signals using a Deep Learning Approach. In: International Conference on Pattern Recognition and Artificial Intelligence, New Jersey.

Documents
Localisation of Drone Controllers from RF Signals using a Deep Learning Approach
[img]
[Download]
[img] PDF
localisation-drone-controllers (12).pdf - Whole Document

3MB
Item Type:Conference or Workshop contribution (Presentation)
Item Status:Live Archive

Abstract

Despite their many uses, small commercial Unmanned Aerial Systems (UASs) or drones pose significant security risks. There is, therefore, a need to find methods of detecting, localising and countering these vehicles. This paper presents work towards autonomously localising drone controllers from the Radio Frequency (RF) signals they emit. An RF sensor array is used to monitor the signal spectrum. A Convolutional Neural Network (CNN) is trained to be able to predict the bearing of the drone controller, relative to the sensor, given its output. The position of the controllers can then be calculated from these bearings, provided that at least two such sensors are deployed a reasonable distance apart. The model is able to achieve a mean absolute error of 3.67° in bearing calculation, which translates into a moderate positional error of 40m at a range of 500m.

Keywords:Deep Learning, Machine Learning, Drone, CNN, Convolution Neural Network, Surveillance, Security
Subjects:H Engineering > H640 Communications Engineering
G Mathematical and Computer Sciences > G420 Networks and Communications
G Mathematical and Computer Sciences > G400 Computer Science
G Mathematical and Computer Sciences > G760 Machine Learning
Divisions:College of Science > School of Computer Science
College of Science > School of Engineering
ID Code:34434
Deposited On:11 Dec 2018 16:09

Repository Staff Only: item control page