This challenge is open to sole innovators, industry, academic and research organisations of all types and sizes. There is no requirement for security clearances.

Solution providers or direct collaboration from countries listed by the UK government under trade sanctions and/or arms embargoes are not eligible for HMGCC Co-Creation challenges.

The UK defence and national security community supports its allies through a range of capabilities, including electronic warfare.

Electronic warfare can include various techniques to jam, misdirect or spoof overthe-air radio frequency (RF) communications and satellite navigation systems.

The first step in disrupting enemy communications is identifying the RF emissions they use.

Current approaches often struggle when operators encounter large numbers of RF emissions from friendly forces, civilian infrastructure, commercial technology and adversary systems all at once.

Traditional digital signal processing, which identifies RF emissions by matching their signals against a library of known examples, is no longer effective given the volume of information available.

HMGCC Co-Creation is looking for technology that can help operators detect and classify enemy RF emitters within a crowded RF environment.

Operators need to identify enemy assets and their RF communication protocols, but face several challenges: high levels of RF congestion, short or infrequent transmissions, variable operating conditions, unknown or changing signal characteristics, and the practical constraints of portable field equipment.

We need:

• reliable detection of target signals in congested RF environments
• fast scanning and signal classification
• technology that works on portable, remote hardware with limited computing power
• a system that can be easily adapted to detect new emitters and different RF environments.

Captain Guarnieri and their team are deployed in a highly contested area. The team specialises in electronic warfare and aims to disrupt enemy capabilities, identifying assets ranging from drone swarms to command and control systems.

They are operating close to a civilian urban area and, near friendly forces, trying to analyse enemy signals. Even with capable sensors collecting data, finding the one emitter they need to worry about is difficult.

Intelligence suggests an enemy command and control system is operating nearby. The target emitter only transmits briefly and unpredictably, making it hard to identify and locate.

Captain Guarnieri uses a portable, lightweight software defined radio (SDR), that runs a machine learning algorithm on a Field Programmable Gate Array (FPGA). The system continuously monitors the spectrum, analyses incoming signals and compares them against known patterns and signatures.

When it detects a transmission, the software immediately alerts the operator, shows the supporting evidence, assigns a confidence score and logs the signal data for further analysis.

With this new capability Captain Guarnieri can cut through the RF noise and focus on just the signals that matter.

The outcome should be a demonstrator delivered after a 12-week project, reaching at least Technology Readiness Level (TRL) 4 (basic technology validation in a lab environment.)

Deliverables:

• a working prototype demonstrated on a commercial software defined radio (SDR) and Field Programmable Gate Array (FPGA) architecture
• a software prototype handed over to the sponsors
• a detailed technical report

Essential requirements:

• techniques used to identify RF classifiers must go beyond traditional digital signal processing
• the solution must show how rapid emitter identification can be completed on a deployed SDR with limited resources
• the solution must cover bandwidths from 1KHz to 1MHz within the SDR and FPGA constraints
• include a roadmap towards operational deployment and future development.

Desirable requirements:

• operation in near real time
• confidence metrics and supporting evidence for classifications or detections
• the ability for government technicians to retrain the model on other data sets

Constraints:

• the sponsors will provide example waveforms, such as FSK, PSK and LoRa, but will not supply real or synthetic training data
• the final deployable model must work on a remote platform without regular internet access or extra computing resources
• the model should run at the edge, on a SDR with a small FPGA System-onChip (SoC), for example 20-80k logic cells, 60-200 DSP blocks, without relying on additional CPU or GPU resources

Not required:

• hardware development is not the primary focus of this challenge
• offensive electronic warfare effects
• a framework for integrating the signal classification tool with other applications
• large-scale fixed infrastructure deployments
• horizon scanning without a working demonstrator

Project duration: 12 weeks

Phase 1 Competition opens: Monday 2 November 2026

Phase 1 competition closes: Monday 23 November 2026

Applicants notified: Monday 7 December 2026

Phase 2 competition opens: Monday 7 December 2026

Phase 2 competition closes: Thursday 7 January 2027