Fiber-optic drones: why RF detection and jamming fail — and what works

A fiber-optic FPV drone replaces its radio link with a spool of optical fiber, typically 10 to 20 kilometres long, that unwinds behind the aircraft as it flies. Control commands travel down the glass to the drone; video comes back the same way. The fiber is thinner than a human hair and weighs little enough that a small quadcopter can carry the spool as payload. The design was not invented to defeat detection — it was adopted because it defeats jamming. In contested airspace a radio-controlled drone loses its link the moment it enters a jammed sector. A fiber-guided one does not, because there is no link to break. Detection immunity came as a side effect, and it is the side effect that matters for site protection.

RF and EM detection works by listening. A scanner monitors the spectrum for the control and video channels a drone uses, identifies the protocol signature, and often takes a bearing on the transmitter. Against a conventional drone this is the earliest and cheapest warning available — a passive receiver can pick up a control link five to ten kilometres out, long before the aircraft is visible or audible. Against a fiber-optic drone the same receiver reports nothing, and it is important to understand why. This is not a question of sensitivity, antenna gain or a better algorithm. The aircraft simply does not radiate. There is no faint signal being missed at the noise floor — there is no signal. No RF scanner on the market, at any price, will ever see one. The same holds for drones flying a pre-programmed autonomous route with the radio switched off, and largely for satellite-linked aircraft whose uplink is narrow and directed skyward.

Electronic warfare — jamming, spoofing, protocol takeover — shares one prerequisite: an exploitable signal. A jammer raises the noise floor so the drone can no longer hear its operator. A spoofer feeds false navigation data. A takeover system speaks the drone's own protocol back to it. Every one of these techniques needs a radio channel to attack. A fiber link runs inside glass, physically isolated from the electromagnetic environment. You can saturate the entire spectrum around the aircraft and the pilot will keep flying with an undisturbed picture. This is why NATO's Allied Command Transformation issued a public request for innovative solutions specifically against fiber-optic FPV drones: the standard electronic answer does not apply to them.

A fiber-optic drone hides one thing only: its radio emissions. Everything else about it remains physically observable. It still has rotors that move air, a body that reflects radar energy, and a shape that a camera can resolve.

Of all the layers, acoustics is the most direct answer to fiber optics, for a simple reason: sound cannot be switched off. The control link can be moved into glass, the route can be pre-programmed, emissions can be reduced to zero. But as long as the aircraft stays airborne, its rotors are chopping air, and that is audible. A rotorcraft produces a recognisable acoustic picture: broadband noise from flow separation plus tonal components at the blade passage frequency and its harmonics. That structure is what separates a drone from wind in foliage, a passing vehicle or a bird — and the same structure lets the AI tell a multirotor from a fixed-wing aircraft and estimate how many rotors it has. A single node is not a microphone but a microphone array. The difference in arrival time across the array's elements yields a bearing to the source. One node gives a bearing; two or more separated nodes intersect their bearings and produce a point in space with an altitude estimate. This is why nodes are not clustered in one place but distributed around the perimeter with overlapping zones — typically four to eight per site, depending on area and terrain.

The uncomfortable conclusion is that RF detection, which is cheap, passive, legal near airports and offers the earliest warning, is exactly the layer that fails first against a determined adversary. It remains the right foundation — most drones over most sites are ordinary radio-controlled aircraft, and nothing else gives warning at ten kilometres. But a perimeter built on RF alone has a documented, publicly discussed gap, and an adversary who has read the same articles knows how to fly through it. The close-in ring is what closes that gap. Acoustic nodes and radar do not care about the control link, and optics confirm what the other two found. The cost of adding them is far lower than the cost of discovering the gap during an incident.

A3O treats detection as layers that cover each other's blind spots rather than as a single sensor. The DD-6 and DD-12 analysers provide the passive RF perimeter at five to ten kilometres. Acoustic nodes, four to eight of them with overlapping zones, form the close-in ring that hears rotors regardless of emissions. Radar adds physical detection of the airframe. An optical turret, cued by whichever layer fired first, confirms the target visually in day or infrared and holds it in auto-track. All four feeds are fused into a single track, so one aircraft appears as one object rather than four uncorrelated alarms. The operator sees which sensors contributed and with what confidence — and, because the response decision stays with a human, decides what happens next.

If your current protection relies on RF detection alone, the gap described here already exists on your site. A vulnerability audit maps where a fiber-guided or radio-silent aircraft could approach unseen, and a short pilot shows what the close-in layers would actually register over your airspace.