Dr Colin Horne - University College London

Radio frequency (RF) signals are pervasive in the modern world, surrounding us all wherever we go. We choose to interact with some of these signals using television receivers, mobile phones, and global positioning system (GPS) receivers, amongst many others, but have little knowledge of what else might be out there. Signal intelligence is the art of gathering information using electronic sensors and is a valuable tool in the understanding of the RF environment. This gives the user visibility of the sensing, communications, and other signals around us.  
 
In today's highly congested and contested RF environment it is critical for both defence and security related activities that we have the capability firstly to detect ‘signals of interest’ amongst the many legitimate background signals, and secondly, to classify their parameters and nature. The sheer volume of intercepted signals is likely to be substantial and ever expanding, requiring radical new ways to enable the detection, identification and mapping of the underlying signal sources. 
 
Dr Horne’s research will investigate highly capable hardware systems for capturing RF signals, along with state-of-the-art processing technologies which can operate in close proximity to the sensor system, to provide the ability to extract accurate and timely signal intelligence. The close proximity or ‘edge-based’ processing allows information to be extracted from the scene without transmitting the high-bandwidth raw data to a central processing hub, thereby reducing signal traffic within the environment under investigation.  
 
The research will leverage recent developments in digital signal processing and machine learning techniques for filtering background signals and analysing the residual environment, highlighting signals of interest. The research programme will culminate in a proof-of-concept demonstrator system that can be fielded in realistic scenarios to meet challenging end user requirements of RF signal detection and classification.