RF drone detection.
Understand the signal. Locate the source.

Find the signal. Understand where it is coming from.

A drone system can emit several kinds of radio traffic: command and control, video, telemetry and Remote ID broadcasts. RF sensing examines the transmissions it supports. Some systems recognize signal characteristics; others extract data from a known protocol or use multiple radio measurements to estimate where a transmitter is located.

The simplest distinction is between reading information inside a signal and measuring the signal itself. Remote ID and supported DJI Drone ID can carry aircraft details and coordinates. Direction finding and TDOA use radio measurements to work out the source’s bearing or position, including situations where useful location data cannot be decoded.

These functions answer different operational questions. Recognizing a likely drone link helps establish that relevant radio activity is present. Decoding an identifier helps distinguish observations. Obtaining aircraft coordinates or locating a transmitter puts the activity into a geographic context.

An aircraft may be non-cooperative with the monitoring system while still emitting a detectable signal. Conversely, a radio-silent aircraft offers no transmission for an RF receiver to acquire. The term “non-cooperative” should therefore be accompanied by the specific signal capability being discussed.

Four ways RF sensing contributes information

Remote ID is an example of protocol decoding: aircraft coordinates arrive inside the broadcast. A receiver can read them without measuring the aircraft’s distance. TDOA and direction-finding systems instead derive information from how the radio wave reaches their receivers. [1] [2]

TDOA: multiple sensors working together to locate a signal

Time difference of arrival, or TDOA, uses multiple sensors at known locations to observe the same radio emission. Because the signal reaches each sensor at a slightly different time, the system can compare those arrival times and calculate where the transmitter is located. It is often described as triangulation; more precisely, it is timing-based multilateration. [2]

The result is a geographic position rather than a direction alone. Precise synchronization, useful sensor spacing and reception of the same signal across the network are essential. TDOA can locate the drone’s transmitter—or a controller transmitter—when the system detects and correctly associates the relevant signal. Accuracy depends on signal quality, network geometry and the propagation environment. [2]

Think of three observers hearing the same sound from different places: it reaches the nearest observer first. TDOA applies that timing principle to radio signals, which travel much faster and require correspondingly precise measurements. The output is a calculated position with uncertainty, rather than an inherently exact point.

In an evaluation, test the edges of the receiving network as well as its center. Move the drone and controller independently. Confirm whether each reported location belongs to the airborne transmitter or the controller; locating one does not automatically locate the other.

RF direction finding: the direction a signal comes from

RF direction finding analyzes a received signal to determine the direction of its source. Bearing-based systems commonly use angle-of-arrival measurements. The result can guide attention toward an emitting drone or controller; bearings collected from different positions can also be combined to estimate a location. [3]

This can be valuable in battlefield and tactical environments, where knowing the direction of an emitter contributes to situational awareness even when the message contents cannot be interpreted. In a city or industrial site, large numbers of unrelated transmitters make signal separation and classification especially important.

Urban structures add a second challenge: multipath. Buildings, vehicles and metal infrastructure can reflect radio energy so that a receiver sees several propagation paths. A strong reflected path may arrive from a different direction than the direct path. The effects should be measured at the actual site, alongside interference and the ability to distinguish drone-related signals from ordinary radio traffic. [3]

TDOA and bearing measurements can also be combined in a hybrid system. Choosing a method is an engineering decision about the environment, required outputs and available installation sites. AirSentinel’s civilian RF deployments use decoded location information from supported signals rather than TDOA or bearing-based location estimates.

Demodulation, decoding and decryption

Demodulation recovers information from the variations in a radio waveform. Decoding interprets that information according to a supported format or protocol. Decryption recovers information protected by encryption. These terms describe different processing steps; detecting or demodulating a signal does not establish that its encrypted contents can be read.

Ordinary Remote ID reception reads a standardized identification broadcast. Manufacturer-link processing is a separate capability. DJI’s historical AeroScope documentation, for example, describes identification and telemetry carried on its drone–controller communication link. That history explains the distinction; it is not a compatibility claim for every current DJI aircraft or firmware release. [4]

Within the AMS family, advanced manufacturer-link acquisition and processing belong to Pro and Pro Max configurations. AMS V10 provides the Remote ID foundation. Model-specific protocol support determines which additional data can be delivered.

Evaluate the usable coverage, not just a range figure

For a facility, usable coverage means detecting the relevant flight at the relevant place and height, then sustaining information long enough for a decision. Test low approaches, parking areas, plant structures and urban corridors, including periods of busy Wi-Fi activity. Repeat flights in different directions and with different supported aircraft.

Keep three problems separate in the acceptance report: a false detection with no corresponding drone, a real drone with an inaccurate location, and a correctly detected flight that triggers an unnecessary alert. Each requires a different improvement—sensing or classification, geolocation, or alert policy.

Test the receiving and delivery chain together. Record first detection, update continuity, source age and the time the integrated platform receives the information. Losing the signal, losing the network connection and delaying an alert should be distinguishable events.

Supported signals. Immediate operational context.

AMS Pro acquires supported DJI O1–O4 OcuSync Drone ID signals and provides aircraft latitude and longitude. Model, serial number and other available fields depend on the aircraft and firmware. AMS Pro Max adds advanced decryption for supported DJI O1–O4 and selected Autel aircraft.

AMS sensors provide local and cloud APIs. Atlas can ingest third-party sensor data and share AMS and fused information with compatible C2 and intelligence platforms. Radar and optical integrations add physical tracking and visual assessment to the RF picture.

AirSentinel has been deployed by customers replacing TDOA- and direction-finding-based systems after experiencing false positives. That customer experience informs our deployment approach; site acceptance testing establishes the performance of a particular installation.

Use the drone detection evaluation guide to define the required coverage, location quality and delivery timing before a field trial.

Common questions