Doppler radar vs RF: two ways to detect a drone
A drone detection radar sees the aircraft; RF drone detection hears its link. Where each one wins, where each one is blind, and why a serious counter-UAS layer runs both.
The two most common first-line sensors in a counter-UAS system work on different signals. A drone detection radar transmits, illuminates the airframe and reads the echo, so it detects the aircraft whether or not it is communicating with anyone. RF drone detection transmits nothing and listens for the control and video link, so it identifies the drone model and often finds the operator, but only while something is radiating. Neither covers the other’s gap, which is why sites that take the threat seriously install both.
How a drone detection radar works
A radar sends out pulses of radio energy and measures what comes back. The delay gives range, the antenna direction gives bearing and elevation, and the Doppler shift of the echo gives the speed of the target towards or away from the radar.
Small drones are a hard target for three reasons, usually summarised as low, slow and small. They fly close to the ground, where returns from terrain, buildings and vegetation are strong. They move slowly, so their Doppler shift is close to that of clutter. And they reflect very little energy: a small quadcopter has a radar cross-section of the order of hundredths of a square metre.
A pulse-Doppler radar deals with this by separating targets from clutter according to velocity, and by looking at micro-Doppler: the rapid modulation that spinning propellers add to the echo. A bird’s wingbeat and a rotor produce different signatures, which is the basis for classifying a track as drone, bird or something else.
The Ga-5000 drone detection radar is an example of a sensor built for this problem. It is an X-band pulse-Doppler radar operating at 9.2 to 9.8 GHz, with published detection of a 0.01 m² target at 5 km or more and reporting of larger targets out to 15 km. A rotating antenna covers 360° in azimuth and 0 to 40° in elevation, with a selectable update cycle of 3 or 6 seconds. Each track is classified from its micro-Doppler signature and reported with position, altitude, course and speed.
What radar is good at
- It needs nothing from the target. A drone on a pre-programmed route, on inertial or visual navigation, or guided over a fibre-optic cable emits no usable radio signal. Radar still sees it.
- It measures position directly. Range, bearing and elevation come from one sensor at one location. The Ga-5000 specification lists range accuracy of 10 m RMS or better and angular accuracy of 0.5° or better in azimuth and elevation.
- It keeps the track when the link goes. If a jammer cuts the control link, or the drone switches to a failsafe mode, an RF sensor loses its signal. The radar track continues.
Where radar struggles
- Line of sight. Radar does not see through buildings or terrain. A drone that approaches below a ridge line or along a street is hidden until it emerges.
- Clutter and birds. Classification reduces false alarms but does not remove the need to manage them. Masking sectors with known clutter helps; the Ga-5000 operator terminal provides suppression zones defined by bearing and range for that purpose.
- Hovering targets. A stationary drone has no body Doppler, so detection relies on the micro-Doppler from its propellers.
- It transmits. A radar can be detected by an adversary with the right receiver, and it needs frequency clearance at the site.
- It does not name the model or find the pilot. Radar classifies a track as a drone. It does not say which product it is or where the operator is standing.
How RF drone detection works
Most drones are in continuous radio contact with a controller. The aircraft sends telemetry and video down; the controller sends commands up. A passive RF sensor is a wideband receiver that scans the bands these links use and compares what it receives with a library of known protocol signatures.
Because each manufacturer’s link has a characteristic structure, a match identifies the make and often the model. Direction-finding antennas give a bearing to the emitter (angle of arrival). Several stations that timestamp the same signal can compute a position from the differences in arrival time (TDOA). For some protocols, the telemetry itself carries the position of the drone and the controller, and the sensor can read it.
The Orbit UAV RF drone detection sensor follows this pattern. It covers 300 MHz to 6 GHz, recognises over 300 drone models by protocol signature including DIY FPV builds and analogue or digital video links, and gives a bearing accurate to 3° RMS or better over 360°. The published detection radius is 3 km in open terrain against a drone transmitting at 0.1 W, with protocol-based positioning of the drone and its remote control to within 2 m. A network of stations adds TDOA positioning, listed at 30 m or better inside the network.
What RF detection is good at
- Early warning. The link is active from the moment the operator powers up, before the drone leaves the ground. An RF sensor can raise the alarm during pre-flight, when there is nothing in the air for a radar to see.
- Identification. Knowing the model tells the security team about likely payload, speed and flight time, and separates a consumer camera drone from a custom FPV build.
- Finding the operator. The controller is an emitter too. Locating it gives security staff or police someone to approach, which often ends the incident faster than dealing with the aircraft.
- Passive operation. The sensor emits nothing. It needs no frequency licence, causes no interference and does not reveal that the site is monitored.
Where RF detection is blind
- Silent drones. An aircraft that flies a stored route with its radio off, or is controlled over fibre, gives an RF sensor nothing to detect.
- Unknown protocols. A link that is not in the library may be detected as energy without being identified. Libraries need updating as new products and custom links appear.
- Position quality from one station. A single direction-finding sensor gives a bearing line, not a point, unless the protocol reveals the position. A fix on an arbitrary emitter needs several stations.
- Range depends on the drone’s transmitter. Detection radius is quoted for a given transmit power. A weaker or more directional link is heard at a shorter distance.
Drone detection radar vs RF detection: side by side
| Drone detection radar | RF drone detection | |
|---|---|---|
| Signal used | Echo from the airframe | Control, telemetry and video link |
| Detects autonomous or fibre-guided drones | Yes | No |
| Detects before take-off | No | Yes, once the link is live |
| Identifies make and model | No, class only (drone, bird, vehicle) | Yes, by protocol signature |
| Locates the operator | No | Yes, when the controller is transmitting |
| Position from one sensor | Range, bearing, elevation | Bearing; full position for known protocols |
| Survives link jamming | Yes | No |
| Emissions | Transmits | Passive |
| Main limitation | Line of sight, clutter | Needs a radiating target |
| Orbit UAV published range | ≥5 km at RCS 0.01 m² | 3 km, open terrain, 0.1 W transmitter |
Why a counter-UAS layer runs both
Each sensor answers a different question. RF says what the drone is and where its operator is. Radar says where the aircraft is now and where it is heading, and continues to say so if the link disappears. Combining them has three practical effects.
Fewer false alarms. A radar track classified as a drone that coincides with an RF detection on the same bearing is very unlikely to be a bird. A radar track with no RF signature is still reported, and is the case that deserves most attention: it may be an autonomous aircraft.
Coverage of the silent threat. Fibre-guided and pre-programmed drones are specifically designed to defeat RF detection and jamming. Radar is the sensor that closes this gap.
A track that survives countermeasures. If the response includes jamming, the RF sensor’s own picture degrades at the moment it is most needed. The radar track is unaffected.
A third sensor completes the chain. Neither radar nor RF shows the operator what the object looks like or what it is carrying. An EO/IR PTZ tracking camera slaved to the radar track provides that visual confirmation; the hand-off is described in our post on slew-to-cue and radar PTZ integration.
Choosing for a specific site
If the budget allows only one sensor, the threat decides. Where the concern is hobbyists and commercial drones straying into controlled airspace, RF detection alone catches most events and finds the pilot. Where the concern includes deliberate intrusion by someone who will take steps to avoid detection, radar is necessary, because that is the scenario in which the drone will not be transmitting.
Site geometry matters as much as the sensor. A radar wants a high, unobstructed position. RF stations want separation from each other for TDOA and distance from strong local emitters. Large or complex sites such as airports, power plants and ports generally need more than one of each; the critical infrastructure page describes how the layers are combined.
FAQ
Can radar detect small drones?
Yes, if it is designed for them. Detecting a small drone requires a radar with enough sensitivity for a very low radar cross-section and Doppler processing that separates slow targets from ground clutter. The published figure for the Ga-5000 is detection of a 0.01 m² target at 5 km or more.
What is the difference between radar and RF drone detection?
Radar is active: it transmits and detects the echo from the aircraft itself. RF detection is passive: it listens for the radio link between the drone and its controller. Radar finds drones that are not transmitting; RF detection identifies the model and locates the operator.
Can RF detection find a drone that is flying autonomously?
No. If the drone is not transmitting and nothing is being transmitted to it, there is no signal for an RF sensor to receive. The same applies to drones controlled over a fibre-optic cable.
How does a radar tell a drone from a bird?
Mainly through micro-Doppler. Rotating propellers add a characteristic modulation to the echo that differs from the pattern of a wingbeat. Track behaviour such as speed, altitude and flight path adds further evidence. Classification is probabilistic, which is one reason to confirm with a second sensor.
Do I need both radar and RF detection?
For a site where the realistic threat is careless or unauthorised hobby flying, RF detection may be sufficient. For a site that has to assume a deliberate, prepared intruder, both are needed, usually with a camera for visual confirmation.