Military tactical datalink emission detection from orbit
Spaceborne wideband RF sensors can detect the presence and approximate location of military tactical datalink emissions such as Link 16 without decrypting them. This page covers what the open literature says about detection physics, geolocation accuracy, revisit limits, and why the application sits in legally and diplomatically sensitive territory.
Sensors
- HawkEye 360 RF constellation: Clusters of three satellites flying in formation use time-difference-of-arrival and frequency-difference-of-arrival (TDOA/FDOA) to geolocate emitters. Published geolocation accuracy is cited by the company as approximately 1 km CEP under favourable geometry, covering VHF through Ku-band including the UHF and L-band ranges where tactical datalinks operate. Revisit at mid-latitudes is multiple times per day across the growing constellation.
- Kleos Space RF constellation: Operates clusters of four satellites for TDOA-based geolocation across HF to UHF. The system is designed for signals of interest in the maritime and defence domains. Geolocation accuracy and revisit figures in the open literature are less precisely published than HawkEye 360's, but the cluster architecture targets sub-kilometre CEP under good satellite geometry.
- Spire Global SDR payload: Spire's LEMUR-2 satellites carry software-defined radio payloads capable of wideband spectrum capture across portions of VHF, UHF and L-band. The SDR approach allows post-collection reconfiguration of the detection window. Spire does not publish geolocation accuracy for RF emitters; the primary capability is spectrum sensing and signal characterisation rather than precise geolocation.
- Generic TDOA/FDOA formation-flying geometry: Any formation of three or more LEO satellites with synchronised clocks and known baselines can attempt TDOA/FDOA geolocation. Accuracy degrades with short baselines, poor satellite elevation angles, and multipath. At LEO altitudes of roughly 500 to 600 km, a 1-microsecond timing error translates to roughly 300 m of range error, setting a practical floor on what clock quality and signal bandwidth can achieve.
What Link 16 looks like from 500 km up
Link 16 is a TDMA waveform operating in the 960 to 1215 MHz L-band, using frequency-hopping across 51 channels at a hop rate of approximately 77,000 hops per second. It is used by NATO and partner-nation aircraft, ships and ground terminals to share situational awareness and fire-control data. The content is encrypted and the waveform is spread across spectrum to resist jamming, but none of that matters to a passive spaceborne receiver trying to detect the emission rather than decode it.
From orbit, the signal appears as a detectable energy burst in the L-band window during each transmission slot. A wideband SDR payload with sufficient sensitivity can register the presence of frequency-hopped energy, characterise the hop pattern and timing structure, and compare those characteristics against the published waveform parameters that appear extensively in open NATO STANAG documentation, academic signal-processing literature, and declassified technical reports. Recognition does not require decryption. The distinction matters legally and operationally: detecting that a Link 16 net is active in a given area is categorically different from reading the messages it carries.
Geolocation: what the physics allows and what it does not
TDOA/FDOA geolocation requires at least three receivers with precisely synchronised clocks and well-known positions. The timing precision needed is severe: L-band signals travel at the speed of light, so a 1-nanosecond clock error introduces roughly 30 cm of range ambiguity. In practice, LEO formation satellites achieve inter-satellite timing synchronisation in the tens-of-nanoseconds range using GNSS-disciplined oscillators, which translates to geolocation errors on the order of hundreds of metres to a few kilometres depending on satellite geometry, signal duration, and signal-to-noise ratio.
Frequency-hopping waveforms like Link 16 complicate TDOA because each hop is very short. The receiver must capture enough hops to accumulate sufficient signal energy for a reliable time-difference measurement. Longer transmissions from ground terminals help; brief airborne transmissions from a fast-moving platform introduce Doppler ambiguity that must be resolved separately. The honest summary: a stationary or slow-moving ground terminal running a Link 16 net is a more tractable target than an F-35 transmitting at 900 knots. Published literature from academic groups working on TDOA of frequency-hopped signals confirms geolocation is feasible but accuracy degrades significantly with short signal duration and poor satellite geometry.
What emission presence tells you, and what it does not
Detecting that a Link 16 net is active in a 1 to 3 km area gives an observer several inferrable facts: military platforms equipped with that specific datalink standard are present or operating nearby, the net is in an active rather than emission-controlled posture, and the approximate time and location of activity can be logged. Over multiple passes, patterns of activation and silence can indicate exercise schedules, operational tempo shifts, or the arrival of new platform types in a theatre.
What it does not give you is message content, platform identity below the level of waveform type, or precise platform count. A single terminal and a hundred terminals produce the same detectable emission signature to a passive receiver. The waveform type constrains the likely platform set to nations and operators that field Link 16 or compatible systems, which is itself useful order-of-battle information, but it is not a substitute for decrypted traffic analysis. Analysts working from emission-presence data alone are doing a form of signals intelligence that has been practised since the Second World War; the satellite delivery mechanism is new, the inferential limits are not.
Revisit, latency, and the gap problem
The fundamental operational constraint is that LEO RF satellites are not persistent. HawkEye 360's published revisit figures suggest multiple passes per day at mid-latitudes across their current constellation, but passes are measured in minutes, not hours. A military unit that enforces strict emission control and transmits only during the gaps between satellite passes, using published or estimated orbital mechanics, can substantially reduce detection probability. This is not a theoretical concern: emission control relative to overhead collection windows has been a standard military practice since satellite reconnaissance became widespread.
Latency from collection to delivered geolocation product varies by operator and tasking arrangement. Commercial RF operators typically quote latency in hours rather than minutes for standard products. For a time-sensitive intelligence application, that is a significant constraint. The revisit and latency figures mean spaceborne RF monitoring is better suited to pattern-of-life analysis over days and weeks than to real-time tactical tracking.
Geopolitical and legal sensitivity
Passive RF collection from orbit occupies an ambiguous position in international law. The Outer Space Treaty of 1967 does not prohibit passive listening from orbit, and the principle of freedom of overflight in space is well established. However, several national legal frameworks restrict the export, re-export, or brokering of systems capable of signals intelligence collection, including the US International Traffic in Arms Regulations and equivalent EU controls. A commercial operator selling RF geolocation services that demonstrably target military waveforms is operating in a space where export licences, end-user certificates, and customer due diligence are not optional formalities.
There is also a distinction between selling a detection capability to a government for use within its own sovereign airspace or territorial waters, and selling it for monitoring third-party military activity. The former is broadly defensible; the latter moves into territory that most commercial RF operators discuss only under non-disclosure. Any procurement of this capability by a government client should involve legal review of both the supplier's export authorisations and the client's own national intelligence law. Satellize, as a non-aligned commercial analytics provider, treats this class of application with the same diligence it applies to any dual-use data product.
Building a monitoring programme around sparse passes
Despite the revisit constraint, a structured spaceborne RF monitoring programme for tactical datalink activity can produce durable intelligence value. The method is pattern-of-life analysis: collect emission events across many passes, build a statistical picture of when and where specific waveform types are active, and flag deviations from the established baseline. A garrison that runs Link 16 nets on a predictable schedule, then suddenly goes silent, is as informative as one that suddenly activates in an unexpected location.
Fusion with other open-source indicators, including SAR imagery of vehicle concentrations, AIS data for naval vessels, and GNSS jamming zone maps from separate sensors, allows analysts to cross-cue between data types and reduce the ambiguity that any single sensor leaves unresolved. Satellize's analytics architecture is built for exactly this kind of multi-source fusion, drawing on open constellations and commercial tasking. If you are scoping a sovereign RF monitoring programme and want to understand what a realistic detection and geolocation specification looks like given current commercial sensor capability, the right starting point is a technical feasibility review against your specific area of interest and target waveform set.
Typical figures
| Frequency coverage (HawkEye 360) | VHF to Ku-band; UHF (300–3000 MHz) and L-band (1–2 GHz) covered |
| Geolocation accuracy (TDOA/FDOA, favourable geometry) | Approximately 1 km CEP published by HawkEye 360; degrades to several km under poor geometry or short signal duration |
| Revisit frequency (mid-latitudes, HawkEye 360) | Multiple passes per day; individual pass duration approximately 5–10 minutes over a fixed point |
| Link 16 waveform frequency range | 960–1215 MHz (L-band), 51-channel frequency-hopping, ~77,000 hops/second (published STANAG 5516) |
| Minimum detectable signal | Dependent on transmitter EIRP, satellite receiver sensitivity, and integration time; not published in open literature for specific military waveforms |
| Product latency (standard commercial delivery) | Typically hours, not minutes; real-time delivery not available from current commercial operators |
| Archive depth | HawkEye 360 commercial archive from 2019 onwards; Kleos from 2020 onwards |
| Delivery formats | Geolocation point files (GeoJSON, KML, shapefile), time-stamped emission logs, API feed |
| Timing precision (inter-satellite, GNSS-disciplined) | Tens of nanoseconds; 1 ns error ≈ 30 cm range ambiguity |
Analytics Satellize can run
| Emission presence map | Waveform classification against published STANAG/open-literature signal parameters; energy detection in target frequency band | GeoJSON point layer of detected emission events with timestamp, frequency band, and waveform-type confidence score |
| Pattern-of-life activity timeline | Time-series aggregation of emission events across multiple passes; statistical baseline construction and deviation flagging | Weekly or monthly activity report with activation frequency, time-of-day distribution, and anomaly flags |
| Geolocation uncertainty polygon | TDOA/FDOA error ellipse propagation using satellite geometry and published timing precision figures | Per-event uncertainty polygon (GIS layer) showing 50% and 90% CEP contours for each detected emission |
| Multi-sensor cross-cue alert | Spatial and temporal fusion of RF emission events with SAR-derived vehicle concentration data and GNSS jamming zone products | Fused alert report correlating RF activity with concurrent imagery or jamming indicators in the same area of interest |
| Emission control gap analysis | Comparison of satellite pass windows against emission event timestamps to estimate probability of detection given a specified emission schedule | Technical briefing note on detection probability as a function of pass geometry and assumed emitter behaviour |
| Waveform type discrimination report | Spectral and temporal feature extraction to distinguish Link 16 from other UHF/L-band waveforms (JTIDS, MIDS, civilian DME) using published waveform characteristics | Classified emission log with waveform-type attribution and confidence interval per event |
Who does the work
We can get this done for you. Satellize runs its own analyst desk and a strong science team. You do not buy a data feed and work out what it means; our people source the imagery, run the analysis described on this page, and hand you the answer with its confidence limits stated. Discuss this requirement.