Integrated Sensing and Communication for FMCW Radar

Radar and communication systems have traditionally operated as separate, independent functions – each requiring its own antenna, RF chain, and processing hardware. On platforms where size, weight, power, and cost are constrained, such as drones, autonomous vehicles and portable defence systems, duplicating this hardware is not always practical. It also places increasing pressure on an already congested electromagnetic spectrum.

The growing demand for connected and intelligent systems means that many platforms now need both sensing and data transmission. Drones need to detect obstacles while receiving navigation commands. Vehicles need to track what is ahead while sharing safety data with the road around them. The question is whether both functions can be combined into a single system, without degrading the performance of either.

• Could a Frequency Modulated Continuous Wave (FMCW) radar transmit data and perform sensing simultaneously, with no compromise to its primary function?

Working in collaboration with University College London (UCL), Plextek developed a novel technique that embeds communication data directly into standard FMCW radar waveforms using frequency hopping, with minimal impact on.

The core innovation lies in what we call the “chirplet” method. A conventional FMCW chirp sweeps continuously across a band of frequencies. Our approach breaks that chirp into discrete sub-band sections – chirplets – and reorders them to encode data. By restricting the hopping so that every frequency band is covered exactly once per chirplet, the full radar bandwidth is preserved. On the receive side, a simple reordering of the in-phase and quadrature (IQ) samples recovers the complete sensing information, allowing the use of conventional FMCW processing with standard Fast Fourier Transforms (FFTs).

This is the key distinction from other frequency-hopped Integrated Sensing and Communication (ISAC) approaches, which require computationally expensive 2D matched filters to recover the radar data. Those methods demand significantly more processing power, making them difficult to implement in real-time systems. Our technique avoids this entirely, adding only minimal computational overhead above what a standard FMCW radar already requires.

The waveform architecture also shares characteristics with LoRa communication signals, giving it a similarly high tolerance for noise, which is an important advantage in congested electromagnetic environments.

We developed a real-time prototype demonstrating simultaneous radar sensing and data communication from a single antenna and RF chain. The technique was validated through simulation and over-the-air experimentation, including tracking a moving person in real time. Further scenarios involving vehicles, drones, and other targets have been validated through modelling and feasibility studies.

The system achieved data rates of several megabits per second while maintaining radar performance. At typical operating parameters, the energy loss introduced by the chirplet technique was negligible – at a fraction of a percent at short ranges – confirming that sensing quality is preserved. The technique was also validated against real-world targets, producing clean Range-Doppler surfaces functionally equivalent to conventional FMCW radar outputs.

This work has significant implications across multiple sectors. In defence, a single ISAC system could simultaneously track hostile drones while sending navigation commands to friendly assets, eliminating the latency that comes from switching between separate radar and communications systems. Positioning, navigation and timing (PNT) nodes could broadcast positioning signals while performing radar surveillance. In civilian applications, automotive radar systems could detect obstacles ahead while exchanging safety data with nearby vehicles, and future 5G/6G infrastructure could combine sensing capability with communications as spectrum becomes more contested.

A journal paper detailing the technique has been accepted by IEEE. This work was carried out in collaboration with UCL and supported by Innovate UK.

• In 2026, this project was recognised with the Technology Company of the Year award at the Cambridge Independent Science and Technology Awards.