Satellite communications interference source location

Uplink interference, whether accidental cross-polarisation or deliberate jamming, can be geolocated by comparing the time and power differences of the same signal arriving at adjacent satellites. This page covers the physics, the published ITU procedures, and the practical limits of the method.

Sensors

Intelsat IntelsatOne Flex interference monitoring network: Intelsat operates a network of beacon monitoring receivers and telemetry channels across its GEO fleet to detect anomalous carrier power on transponders; the system flags interference events and supports TDOA (time-difference-of-arrival) calculations using adjacent satellite pairs separated by as little as 2–4 degrees of orbital arc.
SES network operations centre monitoring: SES uses continuous carrier monitoring across its GEO and MEO fleet to detect uplink interference; adjacent-satellite TDOA and FDOA (frequency-difference-of-arrival) measurements are used operationally to narrow a source to a geographic ellipse, typically hundreds of kilometres wide before ground-truth narrowing.
HawkEye 360 RF geolocation constellation: A LEO cluster constellation (clusters of three satellites in formation) that measures TDOA and FDOA across multiple frequency bands from VHF through Ku-band; published geolocation accuracy is on the order of hundreds of metres to a few kilometres depending on geometry and signal duration; revisit at a given point is roughly 30–90 minutes depending on latitude and cluster configuration.
Kratos Compass and Monics interference monitoring systems: Ground-based carrier monitoring platforms widely deployed by satellite operators; Monics provides continuous per-carrier power and frequency tracking, feeding TDOA workflows when correlated with a second monitoring site or an adjacent satellite's telemetry stream.
ITU-coordinated adjacent satellite TDOA (operator-agnostic method): The ITU Radio Regulations and Resolution 762 establish the procedural framework under which operators share telemetry from adjacent satellites to perform TDOA geolocation; the method requires at least two satellites receiving the interfering signal, with their ephemeris data accurate to better than roughly 50 metres to keep location error below 50 km.

Why the interferer cannot simply hide behind a clean carrier

Every uplink signal that reaches a geostationary satellite also spills onto adjacent satellites in the arc. The arc spacing between commercial GEO satellites is typically 2–4 degrees, which corresponds to a light-travel-time difference of roughly 1–10 microseconds depending on the geometry and the ground station's position. That tiny time difference is the foundation of TDOA geolocation. The interferer does not need to do anything wrong beyond transmitting on the wrong frequency, at the wrong polarisation, or toward the wrong satellite. Physics does not care about intent.

Deliberate jammers face the same constraint. A jammer pointed at a broadcast satellite is simultaneously illuminating its neighbours. The jammer can reduce power to avoid detection, but below a certain threshold the target satellite stops being jammed, which defeats the purpose. There is no practical power level that jams one satellite while being invisible to the adjacent ones, given modern monitoring sensitivity.

The TDOA and FDOA method: what the maths actually requires

TDOA measures the difference in arrival time of the same signal at two satellites. Each TDOA measurement constrains the source to a hyperboloid in three-dimensional space. With two satellite pairs, two hyperboloids intersect along a curve; adding a third pair, or combining TDOA with FDOA (which exploits the Doppler shift difference caused by the source's velocity relative to each satellite), collapses the solution to a point or a small ellipse on the Earth's surface.

In practice, GEO-based TDOA geolocation produces a location ellipse rather than a point. The major axis of the ellipse is typically aligned along the satellite arc direction and can span 50–500 km depending on arc separation, satellite ephemeris accuracy, and signal bandwidth. Narrowband carriers, such as a single SCPC voice channel, are harder to time-stamp precisely than wideband carriers. A 36 MHz transponder bandwidth gives much better TDOA resolution than a 1 MHz carrier. This is an honest and important limit: TDOA alone rarely gives street-level accuracy from GEO. It gives a search area.

LEO RF geolocation constellations such as HawkEye 360 operate differently. Because the satellites are moving, FDOA is large and measurable even over short observation windows. Published HawkEye 360 figures suggest geolocation accuracy of roughly 1–5 km for signals of sufficient duration and power, which is considerably tighter than GEO-only TDOA for the same emitter.

Unintentional interference is the majority of cases

The Satellite Interference Reduction Group (IRG) has documented repeatedly that the large majority of satellite interference incidents are unintentional. Mispointed dishes, misconfigured modems transmitting on the wrong polarisation, and cross-polarisation leakage from poorly maintained uplink systems account for the bulk of operator incident logs. The ITU's Resolution 762 and the associated coordination procedures exist precisely because the problem is chronic rather than exceptional.

Unintentional interference is in some respects easier to geolocate than deliberate jamming. The interfering station is typically transmitting a clean, stable carrier with a known symbol rate, which makes TDOA cross-correlation straightforward. The station is also not trying to avoid detection, so it transmits continuously, giving the monitoring system time to accumulate signal. Deliberate jamming, particularly from a mobile or intermittently transmitting source, is harder to pin down because the signal may be present for only seconds.

From ellipse to address: the ground-truth narrowing process

Once TDOA produces a location ellipse, operators use several secondary methods to narrow the search. Antenna pointing data from the interfering signal's power profile across the satellite beam can constrain the source to a sub-region of the ellipse. If the interferer is using a known carrier type, modulation fingerprinting can link it to a specific equipment manufacturer or network operator. Correlation with licensed earth station databases, maintained under the ITU's Master International Frequency Register, often identifies the culprit station directly, particularly for unintentional cases.

For deliberate jamming with a mobile source, the process is more involved. Successive TDOA measurements as the jammer moves, combined with LEO RF geolocation passes if the signal is present long enough, can track the source's trajectory. In documented cases involving broadcast satellite jamming, the location ellipse has been narrowed to a specific city or military installation through a combination of TDOA, beam-edge power analysis, and open-source corroboration. The method is not infallible, but it is well-established enough that the ITU has codified it in operational guidance.

Honest limits: what the method cannot do

TDOA from GEO pairs does not give precise geolocation for narrowband, low-power, or very short-duration signals. A jammer transmitting for fewer than a few seconds may not produce enough correlated signal for a reliable TDOA measurement. Sources near the satellite arc's sub-satellite point suffer from poor TDOA geometry because the time differences collapse toward zero. And any TDOA result is only as good as the satellite ephemeris data: a 100-metre error in a satellite's known position propagates to roughly a 10–30 km error in the ground location, depending on geometry.

Cloud cover and atmospheric conditions are irrelevant here, which is a genuine advantage over optical methods. But radio-frequency multipath, ionospheric delay variation, and transponder group-delay differences all introduce timing errors that must be calibrated out. Operators with mature monitoring infrastructure have characterised these biases; operators relying on a single incident measurement have not.

Satellize can ingest operator-supplied TDOA ellipse data and correlate it with licensed earth station databases, open-source geospatial layers, and LEO RF geolocation passes from commercial providers to produce a prioritised list of candidate sources, delivered as a structured report with confidence intervals. The workflow is the same class of spatial inference used in the Tonga crop-estimation programme: constrain the hypothesis space with multiple independent data streams, then rank the survivors.

What a practical interference investigation looks like

A typical investigation begins when a network operations centre detects an anomalous carrier on a transponder. The first step is to determine whether the interference is on-satellite (a fault in the satellite's own electronics) or terrestrial (an uplink problem). Beacon monitoring and transponder telemetry usually resolve this within minutes.

If terrestrial, the operator requests TDOA measurements from the adjacent satellite operator, a process formalised under ITU coordination procedures. The resulting ellipse is overlaid on the licensed earth station database. If a match exists, the operator contacts the station directly. If no match exists, or if the interference appears deliberate, the ellipse is passed to a secondary analysis workflow. LEO RF geolocation assets can be tasked if the signal is still active. In documented broadcast satellite jamming cases, the full process from first detection to a credible location attribution has taken anywhere from a few hours to several days, depending on whether the source is stationary and whether adjacent operator cooperation is immediate.

Typical figures

GEO TDOA location ellipse (major axis)	50–500 km, depending on arc separation (typically 2–4°), signal bandwidth, and ephemeris accuracy
LEO RF geolocation accuracy (HawkEye 360)	Approximately 1–5 km, published by operator; dependent on signal duration and power
Minimum signal duration for reliable TDOA	Typically several seconds for wideband carriers; longer for narrowband SCPC signals
Frequency coverage	C-band (3.7–4.2 GHz downlink / 5.925–6.425 GHz uplink), Ku-band (10.7–12.75 GHz / 13.75–14.5 GHz), Ka-band (17.7–21.2 GHz / 27.5–31 GHz); LEO RF assets also cover VHF/UHF
Satellite arc separation required	Minimum ~2° for usable TDOA time difference; wider separation improves accuracy
Ephemeris accuracy required	Better than ~50 m in satellite position to keep ground location error below ~30 km
Detection latency (operator monitoring)	Minutes for carrier detection; hours to days for full TDOA geolocation and attribution
ITU regulatory framework	ITU Radio Regulations Article 15, Resolution 762 (WRC-15/19), and the Master International Frequency Register
Archive depth	Operator-dependent; major operators retain carrier monitoring logs for months to years

Analytics Satellize can run

TDOA ellipse computation and visualisation	Time-difference-of-arrival cross-correlation using operator-supplied telemetry from adjacent satellite pairs, following published ITU TDOA methodology	GIS polygon layer (GeoJSON or shapefile) showing the location ellipse with confidence contours, delivered within hours of data receipt
Licensed earth station candidate matching	Spatial intersection of the TDOA ellipse with the ITU Master International Frequency Register and national licensing databases	Ranked list of candidate earth stations with licence details, operator contacts, and match confidence scores, in PDF and structured CSV
LEO RF geolocation tasking and fusion	TDOA/FDOA geolocation from commercial LEO RF constellation passes (e.g. HawkEye 360), fused with GEO TDOA ellipse to reduce the candidate area	Fused location estimate with reduced uncertainty bounds, delivered as a GIS point feature with error radius and pass metadata
Interference event timeline reconstruction	Time-series analysis of carrier monitoring logs to reconstruct transmission schedule, power ramp-up, and frequency drift, consistent with published signal characterisation methods	Annotated timeline report showing event start, duration, intermittency pattern, and inferred transmitter behaviour
Modulation and equipment fingerprinting	Spectral analysis of captured carrier to identify symbol rate, roll-off factor, and modulation type, cross-referenced against known equipment signatures in published technical literature	Equipment type assessment with confidence level, included in the investigation report
Recurring interference pattern monitoring	Automated correlation of new interference events against a historical database of prior TDOA ellipses and attributed sources, flagging probable repeat offenders	Standing alert feed with event-to-prior-case linkage scores, updated per new incident

Sources

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.