Drones and the spectrum on the modern battlefield

In a few short years the small drone — from the consumer quadcopter to the loitering munition — has become a central actor in conflicts. And all of its effectiveness rests on the radio spectrum: a drone that can neither receive commands, nor send back video, nor locate itself is just a toy. That is why counter-drone work is, first and foremost, a matter of electronic warfare.

Educational note: it describes publicly known physical principles (which bands, how detection works, why jamming does or doesn't work). No operational instructions. Transmitting/jamming is forbidden to civilians (Légal & sécurité).

A drone's three radio links

A remotely piloted drone relies on three radio links, each on identifiable bands:

Command & control (C2): uplink, pilot to drone. Often on 2.4 GHz and 5.8 GHz (the same ISM bands as your WiFi), using frequency hopping to resist jamming.
Video downlink: drone to pilot. Wideband (several MHz) — this is the wideband signature you learn to spot in the Capstone mission. Racing FPV drones often use analogue 5.8 GHz video; consumer drones a digital link (OcuSync-type).
Navigation (GNSS): the drone listens to GPS/GLONASS/Galileo to position itself and hold a course. A very weak signal at ground level, hence vulnerable to jamming and spoofing.

Cutting one of these links is often enough to neutralise the drone — or to trigger its safety behaviour (return to home, land, hover).

Detecting a drone by radio

Passive detection (receive-only, like an SDR) looks for the signatures of these links:

Wideband energy at 2.4 / 5.8 GHz: the video downlink occupies a lot of spectrum — that is exactly the drone mission's objective (detecting an emission ≥ 5 MHz).
Frequency-hopping pattern: the C2 link hops at a recognisable cadence, different from household WiFi.
Remote ID: in many countries drones must broadcast in the clear (often over WiFi/Bluetooth) an identifier and the position of both the drone and the pilot. It's a legally decodable signal, very useful for cooperative detection.
Model fingerprint: bandwidth, frequencies and protocol often let you identify the drone's type.

Radio detection has a major advantage: it is passive and silent, and reaches beyond line of sight (the drone gives itself away before it's visible). It combines with radar, acoustics and optics to make the alert reliable.

Neutralising: why it's hard

Countermeasures target the three links:

C2 jamming → the drone loses its orders and triggers its failsafe.
GNSS jamming/spoofing → the drone loses its position; spoofing can even make it drift.
Capture/takeover → exploiting a protocol flaw (increasingly rare, as links are encrypted and agile).

But the adversary answers with agility (electronic protection): band hopping, resistant waveforms, and above all autonomous drones guided by camera and onboard AI that no longer need a radio link once launched — nothing to jam. That's the current frontier: when the drone stops talking, classic EW loses its grip, and the fight shifts to optics, acoustics and the kinetic interceptor.

The race under way

Recent evolution comes down to a few trends:

ISM band saturation: so many drones and jammers that 2.4 and 5.8 GHz become a permanent electromagnetic battlefield.
Area GNSS jamming: whole regions where civilian GPS is unusable, with collateral effects on aviation and maritime navigation.
Fibre-optic drones: tethered to the pilot by a wire several kilometres long — no radio emission at all, hence undetectable and unjammable by EW.
Autonomy and swarms: AI guidance, optical terminal targeting, swarm coordination — shifting value from the radio link to onboard computation.

For a curious civilian, the concrete entry point remains the drone mission and listening to 2.4 GHz: there, in miniature and entirely legally, you see the same wideband-signature physics that structures this whole field.