Drone Detection Systems: RF, Radar, Acoustic and Optical Compared

Every drone you build flies in airspace that someone increasingly wants to watch — airports, prisons, stadiums, power plants, borders and battlefields all now run counter-UAS programs. Whether you're building drones or defending against them, understanding how detection works is fast becoming core knowledge. This guide maps the four sensing modalities and the honest trade-off that runs through all of them: every sensor has a blind spot the others cover.

The counter-UAS chain: detect, track, identify, mitigate

Detection is one link in a chain. A useful system must detect that something is there, track where it is and where it's going, identify whether it's a threat (and ideally which model, and who's flying it), and only then support a decision to mitigate. Mitigation — jamming, spoofing, nets, kinetic — is heavily regulated and often illegal for civilians; this guide is about the sensing half, which is legal to understand and mostly legal to deploy (receive-only). Keep that legal line firmly in mind: detecting is not defeating.

The four modalities at a glance

Modality	Detects by	Typical range (small UAS)	Kills it	Blind spot
RF	Listening for control/video radio links	1–5 km+	Low cost, early warning, can locate pilot	Autonomous/silent drones; RF-noisy cities
Radar	Reflecting radio waves off the airframe	~0.3–several km	Works day/night, all weather, any drone	Small/slow/low targets hide in clutter; birds
Acoustic	Hearing propeller/motor noise	~300–500 m	Passive, cheap, catches silent drones low down	Short range; wind and urban noise
Optical (EO/IR)	Seeing the drone with cameras + AI	Detect ~0.3–2 km, ID closer	Visual confirmation and identification	Needs a cue; fog, night, sun, occlusion

RF: the cheap first tripwire

Most drones constantly emit — a control link, a video downlink, sometimes Wi-Fi or telemetry. RF sensors listen for these signatures, and many commercial systems carry a library of drone protocols so they can name the model and even decode Remote ID. Two receivers give direction finding; three or more give a position by triangulation. It's the best value early-warning layer — and useless against a drone flying a GPS mission in radio silence. Full detail in the RF drone detection guide.

Radar: sees anything, struggles with the small stuff

Radar transmits and measures reflections, so it doesn't care whether the drone emits anything. The problem is physics: a 500 g quadcopter has a tiny radar cross-section, flies slowly and low, and hides in "clutter" — buildings, trees, and especially birds, which look almost identical. Modern counter-UAS radars use micro-Doppler (the flicker of spinning propellers) and machine learning to separate drones from birds, but small-slow-low remains the hard case. Radar is powerful and expensive, and unlike passive sensors it emits, so deploying it touches spectrum rules.

Acoustic: short range, but it hears the silent ones

Propellers and motors make a distinctive whine. A microphone array can detect and roughly localise that sound out to a few hundred metres, passively and cheaply — and crucially it works against autonomous drones that RF can't see. The catch is range and noise: wind, traffic and urban din swamp the signal, so acoustic shines as a close-in, low-altitude layer, especially around sensitive points like prison walls.

Optical: the sensor that identifies

Cameras — visible and thermal — with AI object detection are how a system goes from "something is out there" to "that is a DJI Mavic carrying a payload." Optical usually works as a slave: another sensor cues it where to look, then a pan-tilt-zoom camera confirms and identifies, and two cameras can triangulate a 3D position. It's the confirmation and evidence layer, limited by line of sight, weather, darkness and sun glare.

Choosing a starting layer

1. Define the site and threat. A prison worries about small drops from consumer drones low over a wall (acoustic + optical); an airport worries about anything in approach corridors (radar + RF); critical infrastructure wants layered coverage.
2. Start passive. RF and acoustic are receive-only, cheaper and legally simpler than radar (which emits) or any mitigation (mostly prohibited for civilians).
3. Plan for fusion from day one. Two sensors that don't talk to each other are two sets of false alarms; the value is in combining them — see the sensor fusion guide.
4. Respect the law. Detection is broadly permissible; jamming, spoofing and interception are tightly restricted. Know the line before you buy.

Frequently asked questions

What is the best way to detect a drone?

There is no single best sensor — each has a blind spot. RF detection is cheap and gives early warning but misses autonomous drones; radar sees any moving object but struggles with small, slow, low targets; acoustic works at short range and low altitude; optical confirms and identifies but needs a cue and good visibility. Reliable systems fuse two or more.

How far away can a drone be detected?

Very roughly: RF detection 1–5 km (sometimes more) depending on the drone's emissions, radar hundreds of metres to several km for small UAS, acoustic 300–500 m, optical from a few hundred metres (detection) with identification much closer. Real ranges depend heavily on the drone, the environment and the sensor.

Can you detect an autonomous drone with no radio link?

Not with RF detection, which listens for the control and video links. A drone flying a pre-programmed GPS mission in radio silence is invisible to RF sensors — which is exactly why radar, acoustic and optical sensing, combined by sensor fusion, matter for a complete picture.