Every stealth aircraft ever built shares the same underlying bet: that its shape and coating can defeat radar operating within a known, narrow slice of the radio spectrum. That bet has held for decades because conventional radar generates its signals electronically, and electronic components are physically limited in how much bandwidth they can produce, process, and search across at once. A new class of sensor now threatens to break that bet entirely, not by building a bigger or more powerful version of existing radar, but by replacing the electronics at its core with light.
Photonic radar uses lasers and optical components, rather than electronic oscillators and copper circuitry, to generate and process radar signals. The shift sounds modest. Its implications for airborne platforms, where every kilogram of payload and every gap in stealth coverage carries outsized consequences, are not.
The Electronic Bottleneck
Conventional radar has always faced a hard physical constraint. The electronic oscillators that generate its signals, and the copper circuits that carry them, can only span a limited range of frequencies before signal quality degrades, and they generate a noise floor, background electrical interference, that makes weak returns from small or distant targets difficult to separate from clutter. That bandwidth limit is precisely what stealth designers exploit. Radar-absorbent coatings and faceted airframes on aircraft like the F-35 or China’s J-20 are tuned to specific, well-understood frequency bands, the L, S, and X bands most commonly used by search and fire-control radar. A stealth aircraft does not need to defeat every possible radar frequency. It only needs to defeat the ones a given radar can actually generate.
Photonic radar attacks that constraint at its source. Because light itself carries a vastly higher native frequency than radio waves, an optical signal can be modulated across an extremely wide range, including into the millimetre-wave and terahertz bands, without the bandwidth ceiling that limits electronic generation. Researchers at the University of Sydney’s Institute of Photonics and Optical Science, whose work on photonic radar signal generation has been backed by both the Australian Research Council and the U.S. Air Force, have demonstrated exactly this advantage in published research, generating radar waveforms with more than 30 gigahertz of tunable bandwidth and, in later work, sharply reduced phase noise compared with conventional electronic signal generation. Because the entire signal chain runs through light rather than electrical current, the resulting system is also largely immune to electromagnetic interference and to the radio-frequency jamming that adversaries use to blind conventional sensors, since there is no electrical circuit for a jammer’s noise to couple into in the first place.
Unmasking Stealth From the Air
The most consequential application of this bandwidth advantage is detection range against low-observable aircraft, and it is worth understanding the underlying physics rather than treating it as a marketing claim. Radar performance follows the radar equation, which ties maximum detection range to transmitted power, antenna gain, a target’s radar cross-section, and the receiver’s ability to pull a weak return out of background noise. A stealth aircraft’s entire design exists to minimize that radar cross-section against the specific bands a defender is known to operate in.
A useful way to picture the effect is to compare how a standard electronic radar and a photonic radar handle two very different airborne targets. Against a small consumer-grade drone with a radar cross-section of roughly 0.01 square metres, a conventional X-band radar typically struggles to hold a reliable track beyond about 5 kilometres, since the drone’s weak reflected signal gets lost in the receiver’s own thermal noise. A photonic radar operating at equivalent power can push that detection range out to roughly 12 kilometres, because its far more stable optical signal generation lowers the receiver’s noise floor substantially, letting a weak return stand out more clearly against the background. Against a genuinely low-observable aircraft, whose radar cross-section is reduced by an order of magnitude further still, the gap widens even more. A conventional radar tuned to specific bands may see its usable detection range fall below 15 kilometres, while a photonic system sweeping simultaneously across a much wider slice of spectrum, including bands the stealth coating was never optimized to defeat, can extend reliable detection out to 45 kilometres or more, multiplying the warning time available to a defending force. These figures are illustrative rather than the outcome of a single published field trial, but they follow directly from the radar equation and from the documented bandwidth and noise-floor advantages photonic systems have already demonstrated in laboratory and prototype testing.
That combination, wider bandwidth defeating narrowband-optimized coatings and a lower noise floor pulling out weaker returns, is why defense researchers describe photonic radar as a genuine threat to the stealth paradigm rather than an incremental sensor upgrade.
Inside the Signal Chain
Three components sit at the heart of any defense-grade photonic radar, and each replaces a specific electronic part with an optical equivalent built to solve the same problem more effectively.
The system’s timing reference is a high-coherence laser rather than a crystal oscillator, typically operating with a linewidth in the kilohertz range to achieve extreme phase stability. Defense-grade designs often use a master-slave configuration, in which a highly stable master laser injection-locks a higher-power slave laser, producing a signal with very low phase noise. That stability is what allows the radar to resolve slow-moving targets, a drone crawling along a treeline, for instance, against background clutter that would otherwise mask the return entirely.
Converting an electronic radar signal into the optical domain requires a Mach-Zehnder Modulator (MZM), a device that splits a beam of light into two paths, applies a voltage to one, and recombines them to modulate the light’s intensity. Defense applications typically use Dual-Parallel MZMs, which support the more complex I/Q modulation needed to generate chirped waveforms, signals that sweep smoothly across a frequency range, essential for measuring a target’s range and velocity simultaneously with high precision.
The final stage converts the optical signal back into a microwave signal for transmission or processing, and here defense systems rely on Uni-Traveling-Carrier Photodiodes (UTC-PDs) rather than standard photodiodes. Conventional photodiodes rely on both slow-moving positive charge carriers and fast electrons, which limits how much power they can handle before the signal distorts. UTC-PDs use only electrons as active carriers, letting them operate at far higher power levels and frequencies, up to the terahertz range, which translates directly into longer detection range and sharper resolution. These are frequently paired with balanced photodetectors, which use two matched photodiodes to cancel out noise common to both signals, effectively doubling signal strength relative to the background.
Why Weight Matters More in the Air Than Anywhere Else
Every advantage described above matters more on an airborne platform than almost anywhere else a radar might be deployed, because airborne platforms are uniquely punished for excess weight. Future Military Technologies has covered how defense electronics manufacturers approach the SWaP problem, Size, Weight, and Power, as one of the defining engineering constraints on any system destined for a drone, aircraft, or missile, and radar is no exception.
Conventional electronic radar distributes its high-frequency signals from a central processor to antenna elements along the wings and fuselage using thick, heavily shielded copper coaxial cable, since copper carrying radio-frequency current requires dense insulation to prevent electromagnetic interference from leaking into the rest of the aircraft’s systems. A representative reconnaissance drone radar payload illustrates the scale of that burden well. Heavy-gauge copper cabling and its shielding can account for roughly 30 kilograms of an 85 kilogram total system, with a further 45 kilograms consumed by the central RF processing unit and the aluminium heatsinks needed to dissipate heat from inefficient electronic components, leaving around 10 kilograms for the antenna elements themselves.
Replacing that copper backbone with single-mode fiber-optic cable changes the calculation substantially, since fiber carries light rather than electrical current and needs no heavy shielding against interference. In the same illustrative payload, that 30 kilograms of copper cabling shrinks to roughly 2 kilograms of fiber, and because signal generation and processing occur on lightweight photonic chips rather than power-hungry circuit boards, the central processing unit can shed its heatsinks and drop from around 45 kilograms to 15 kilograms. Combined with the unchanged antenna weight, a photonic radar payload built to the same specification can come in around 27 kilograms, close to 58 kilograms lighter than its electronic equivalent, freeing that weight for additional fuel, extended loiter time, or other mission equipment on a medium-altitude, long-endurance UAV where every kilogram is contested space.
From Passive Sensor to Cognitive Radar
The most forward-looking application of photonic radar pairs its optical reconfigurability with artificial intelligence, producing what researchers call cognitive radar, a system that actively adapts its own physical operating parameters in real time rather than simply detecting and reporting.
Because photonic components can be reconfigured across a wide bandwidth by adjusting the properties of light itself, rather than being locked into the rigid limits of analog circuitry, an AI algorithm monitoring the electromagnetic spectrum can respond to a jamming attempt within microseconds, instructing the system’s optical wave shaper to shift the Mach-Zehnder Modulator’s bias voltage and hop the radar’s operating frequency to a clear part of the spectrum, a frequency change that can execute in under a nanosecond. The same closed loop can manage power and signature discipline on a long patrol, lowering the laser’s pulse repetition frequency to conserve energy and reduce thermal signature while scanning empty airspace, then instantly retuning the photodiodes and narrowing the beam into a high-energy tracking spike the moment a genuine threat appears. If that target attempts to escape detection by dropping into ground clutter, the same AI layer can switch the waveform from a simple linear chirp to a more complex phase-coded sequence to sharpen contrast against the noisy background. This kind of real-time, AI-driven sensor adaptation sits alongside the broader wave of artificial intelligence entering military systems, extending machine-speed decision-making from the battle-management layer down into the physical behaviour of the sensor itself.
Who Is Actually Building This
Photonic radar has moved well beyond laboratory theory, though most programmes remain closer to prototype than fielded system.
In the United States, the Office of Naval Research awarded RTX Raytheon a $20.7 million, three-year contract for a project called Birdseye Yonder (BEYOND), tasking the company with building and demonstrating two prototype wideband passive photonic radio-frequency sensor systems, work being carried out across facilities in Tewksbury, Massachusetts, El Segundo, California, and Arlington, Virginia, with completion expected by November 2026. India’s Defence Research and Development Organisation, through its Electronics and Radar Development Establishment in Bengaluru, has pursued the most publicly documented programme to date, successfully imaging small objects as fine as 3 by 4 centimetres on a rotating test plate as early as 2022, and is now working toward trials that move the system from stationary testbeds onto UAVs and naval vessels, intended eventually to complement or replace conventional AESA radar on fighter aircraft and ground-based air defence. A DRDO official described the technology as offering faster, more precise detection “even in contested environments,” a claim consistent with the jamming resistance built into the underlying architecture.
On the academic side, Professor Benjamin Eggleton’s team at the University of Sydney’s Institute of Photonics and Optical Science, working with funding from the U.S. Air Force and the Australian Research Council, has published a steady line of results pushing photonic radar toward genuine miniaturisation, including a fully integrated photonic chip demonstrated in December 2023 offering 15 gigahertz of tunable bandwidth with spectral resolution down to 37 megahertz, a step explicitly aimed, according to one of the paper’s co-authors, at compact RF photonic filters suited to air and spaceborne payloads. That trajectory, from rack-mounted laboratory demonstrations toward chip-scale integration, is what will ultimately determine whether photonic radar can be deployed across large numbers of small drones rather than a handful of expensive flagship platforms. It is worth noting that photonic radar is a distinct technology from quantum radar, a separate research area that exploits the quantum properties of individual photons rather than classical light, and the two are sometimes conflated in casual coverage despite resting on very different physics.
What Still Stands in the Way
None of this amounts to a finished, fielded replacement for conventional airborne radar yet. Raytheon’s BEYOND programme is explicitly a passive sensing demonstration for geolocation and signals intelligence rather than a general-purpose active imaging radar. DRDO’s system has proven the underlying concept on a stationary bench but has not yet completed the harder engineering step of surviving and performing reliably aboard an actual aircraft in flight. The University of Sydney’s own published chip work describes itself candidly as a pilot study laying groundwork for future integration, not a deployable sensor.
The remaining obstacles are as much industrial as scientific: ruggedising delicate optical components for the vibration, temperature swings, and shock loads of military flight, manufacturing photonic chips at a cost and yield that make sense for large drone fleets rather than single prototype units, and building the supply chain of specialised lasers, modulators, and photodiodes that defense-grade systems will require at volume. Those are solvable engineering problems, and the pace of investment from the U.S. Navy, India’s DRDO, and university research groups backed by military funding suggests none of the major programmes view them as fundamental barriers. But they are the reason photonic radar remains, for now, a rapidly maturing capability rather than something already carried into contested airspace.
Everything depends on physics
Stealth technology has spent four decades winning its arms race against radar by exploiting a narrow, predictable slice of the electromagnetic spectrum that electronic radar was physically bound to operate within. Photonic radar does not attack stealth coatings or airframe shaping directly. It attacks the assumption underneath them, that a defender’s sensor can only ever search a limited band at a time, and it does so while cutting the weight penalty that has historically made powerful radar difficult to fit onto small, endurance-constrained airborne platforms.
Whether that combination arrives on operational aircraft within the next few years or the next decade will depend less on the physics, which multiple independent research programmes have already demonstrated works, and more on the unglamorous work of ruggedisation, manufacturing scale, and integration testing that separates a laboratory result from a combat-ready sensor. For an industry that has spent a generation building stealth aircraft around the limits of conventional radar, that is a genuinely uncomfortable position to be in.
