Passive bistatic radar detection using LEO satellite signals of opportunity
Passive bistatic radar uses LEO satellite downlinks as uncooperative illuminators to detect aircraft, ships and ground targets without emitting a single watt. The technique is experimentally promising but operationally immature, and the signal-to-noise arithmetic is brutal.
Sensors
- GNSS constellations (GPS L1/L2, Galileo E1/E5, GLONASS L1/L2): Continuous global illumination at 1.2–1.6 GHz with effective isotropic radiated power around 27–50 W per satellite. The forward-scatter geometry is well studied for aircraft detection; bistatic cross-section peaks near the specular angle. Coherent integration over seconds can recover targets, but the low EIRP makes direct-path cancellation critical.
- Iridium NEXT (66 satellites, L-band ~1.6 GHz): Iridium NEXT downlinks have been used experimentally as LEO illuminators. The higher Doppler rate relative to GNSS (orbital period ~100 minutes, ground speed ~7.5 km/s) provides better range-Doppler separation for moving targets. Published experiments have demonstrated detection of aircraft at ranges of tens of kilometres under controlled conditions.
- CYGNSS (8 microsatellites, GPS L1 C/A reflectometry): Designed for ocean surface wind retrieval, CYGNSS is the closest operational analogue to space-borne passive bistatic radar. It records GPS signals reflected from the sea surface with a delay-Doppler mapping receiver. Spatial resolution of individual delay-Doppler maps is roughly 25 km, which is useful for oceanographic products but far too coarse for vessel or aircraft detection.
- TechDemoSat-1 (UK, GNSS-R payload, 2014–2018): TechDemoSat-1 carried a space-borne GNSS reflectometry receiver that produced published datasets on ocean roughness and soil moisture. Its data are used in academic literature as a reference for what a low-cost space-borne passive receiver can actually measure. Target detection was not its mission, but it demonstrated the noise floor a space-borne bistatic receiver faces.
- Spire Global GNSS-R payload (Lemur-2 constellation): Spire operates GNSS-R receivers on its Lemur-2 satellites, producing atmospheric and ocean surface data commercially. The payload architecture is public. Researchers have discussed whether similar receivers could be repurposed for target detection, but no operational detection product exists. Revisit at any given point is irregular and depends on constellation geometry.
What passive bistatic radar actually does
Conventional radar transmits a pulse and listens for its echo. Passive bistatic radar does neither. It uses signals already present in the environment, here the downlinks of LEO or GNSS satellites, as unintentional illuminators. A ground receiver, or a space-borne receiver, records two things simultaneously: the direct-path signal from the satellite and the faint echo scattered off a target. Cross-correlating the two reveals the bistatic range and Doppler shift of anything that moved between transmitter and receiver.
The geometry matters enormously. GNSS satellites at medium Earth orbit produce a near-continuous illumination field, but their signals are extraordinarily weak at the surface, around minus 130 dBm in a standard antenna. Aircraft and ships scatter only a tiny fraction of that. The bistatic radar cross-section of a commercial aircraft in the forward-scatter geometry can be large, occasionally exceeding the monostatic value, but only within a narrow angular cone around the specular direction. Outside that cone, detection becomes very difficult without long coherent integration times.
Ground-based detection: what the published record shows
The most credible experimental results come from ground-based receivers using GNSS illuminators. Researchers at University College London and elsewhere have demonstrated detection of commercial aircraft at ranges of 50–200 km using GPS L1 signals and coherent integration over several seconds. The Doppler signature of a crossing aircraft is distinctive: a smooth S-curve in the range-Doppler map as the target's bistatic velocity changes. Ships have been detected at shorter ranges in calm sea conditions, where the clutter environment is more manageable.
Iridium illuminators offer a practical advantage. Because Iridium satellites move at roughly 7.5 km/s relative to the ground, the Doppler shift they impose on a stationary target is large and time-varying. This makes it easier to separate target echoes from direct-path interference and ground clutter. Published work by Griffiths, Baker and colleagues in IEEE Aerospace and Electronic Systems has outlined the theoretical framework and some experimental validations. The results are real. They are also laboratory-grade: achieved with purpose-built receivers, careful antenna placement, and post-processing that is not yet compatible with real-time operational systems.
The space-borne inversion problem is harder than it looks
Detecting ground or maritime targets from a space-borne passive receiver reverses the geometry. Now the receiver is moving at orbital velocity, the target is on or near the surface, and the illuminating satellite may be a third body in a different orbit. The direct-path signal arrives from one direction; the scattered echo arrives from another, Doppler-shifted by the combined motion of the satellite and the target.
CYGNSS has shown that space-borne GNSS-R receivers can measure ocean surface roughness at roughly 25 km resolution with good revisit. Detecting a ship, which presents a bistatic cross-section orders of magnitude smaller than a sea-state cell, is a categorically different problem. The signal excess over the background is tiny. Published analyses suggest that detecting a large vessel (tens of thousands of gross tonnes) in calm conditions might be feasible with long coherent integration, but smaller vessels and aircraft remain below the noise floor of current space-borne receivers. No operational space-borne passive radar detection system for maritime or air targets exists as of the public record.
TechDemoSat-1 data confirmed the noise floor problem directly. The satellite's GNSS-R receiver was sensitive enough for oceanographic work but not for target detection. This is not a failure of the concept; it is an honest statement of where the technology sits.
Signal-to-noise: the arithmetic that governs everything
The bistatic radar equation for a passive system is unforgiving. Received signal power scales with transmitter EIRP, target bistatic cross-section, and the product of transmitter-to-target and target-to-receiver path losses. GNSS EIRP is fixed by design at values optimised for navigation, not illumination. Path losses at 1.5 GHz over hundreds of kilometres are large. The target cross-section for a ship or aircraft is small relative to the clutter background, especially at sea.
Coherent integration over longer dwell times improves the signal-to-noise ratio, but only if the target's motion can be compensated accurately. A ship moving at 15 knots introduces a bistatic Doppler shift of a few hertz at L-band, which is detectable in principle but requires precise knowledge of the receiver's own position and velocity. For a space-borne receiver moving at 7.5 km/s, the Doppler budget is dominated by the satellite's own motion, and separating target-induced Doppler from platform-induced Doppler demands very precise orbit determination. Errors of even a few centimetres per second in the receiver's velocity translate into false Doppler signatures.
Where the technique is credibly useful now, and where it is not
Ground-based passive bistatic radar using GNSS or Iridium illuminators is a credible, if niche, surveillance technique for wide-area aircraft detection in low-clutter environments. It is attractive for sovereign programmes because it requires no transmit licence, no spectrum allocation, and no high-power infrastructure. A country with access to the sky has access to the illuminators. The receiver hardware is relatively inexpensive. The signal processing is the hard part.
Maritime detection from the ground is harder. Sea clutter at L-band is significant in anything above sea state 2, and ships produce weaker forward-scatter signatures than aircraft. Space-borne detection of individual vessels or aircraft is, at present, beyond the demonstrated capability of any published system. Researchers are exploring higher-EIRP LEO constellations such as Starlink as potential illuminators; the Ku-band and Ka-band downlinks offer more power and shorter wavelengths that may improve cross-section for metallic targets, but published experimental results using Starlink as a bistatic illuminator are very early and not yet operationally validated.
What an honest procurement conversation looks like
Any vendor offering operational space-borne passive bistatic radar detection of ships or aircraft today is ahead of the published science. The technology readiness level for space-borne target detection sits at TRL 2–3 by standard ESA or NASA definitions. Ground-based systems using GNSS illuminators are closer to TRL 5–6 for aircraft detection in controlled geometries.
A realistic near-term application is ground-based wide-area aircraft surveillance as a complement to ADS-B, particularly in regions where ADS-B coverage is sparse and the cost of a conventional radar is prohibitive. The system provides no identification, only detection and track. Fusion with AIS or ADS-B data is necessary to assign identity to a detected track. Cloud is irrelevant at L-band, which is one genuine advantage over optical surveillance. Day-night operation is another. But the latency between detection and a usable track can be minutes rather than seconds, depending on integration time and processing architecture.
Typical figures
| Illuminator frequency (GNSS) | 1.164–1.610 GHz (L-band); GPS L1 at 1575.42 MHz most commonly used |
| Illuminator EIRP (GPS) | Approximately 27–50 W effective isotropic radiated power per satellite; very low by radar standards |
| Bistatic detection range (ground-based, aircraft) | 50–200 km demonstrated experimentally under controlled conditions; operational range uncertain |
| Coherent integration time required | Typically 1–10 seconds for aircraft; longer for ships; limits real-time track update rate |
| Space-borne receiver spatial resolution (GNSS-R, oceanographic) | ~25 km for CYGNSS delay-Doppler maps; not suitable for individual vessel detection |
| Technology readiness level (space-borne target detection) | TRL 2–3; no operational system demonstrated in peer-reviewed literature |
| Technology readiness level (ground-based aircraft detection) | TRL 5–6 for controlled experimental geometries; not yet operational at scale |
| Minimum detectable target (ground-based, GNSS illuminator) | Large aircraft (high bistatic RCS near specular) demonstrated; small aircraft and ships remain marginal |
| All-weather capability | L-band penetrates cloud and rain with negligible attenuation; no optical dependency |
| Output product | Bistatic range-Doppler map; track (position and velocity vector, no identity); requires AIS or ADS-B fusion for identification |
Analytics Satellize can run
| Feasibility assessment for ground-based passive radar network | Bistatic radar equation modelling using published GNSS EIRP values, terrain masking analysis, and published aircraft RCS data | Technical report with coverage maps, predicted detection probability curves, and receiver siting recommendations |
| Illuminator geometry analysis | Orbital propagation of GNSS and LEO constellations (Iridium, candidate Starlink geometries) to compute bistatic coverage and Doppler availability over a defined area of interest | Time-series coverage plots and bistatic geometry tables in PDF and GIS layer formats |
| Signal-to-noise budget for target class | Bistatic radar equation applied to client-specified target (vessel class, aircraft type) using published cross-section estimates and atmospheric propagation models | Parametric SNR report with sensitivity to integration time, receiver gain and clutter assumptions |
| Literature synthesis on LEO illuminator passive radar | Systematic review of peer-reviewed publications (IEEE AES, Remote Sensing MDPI, IET Radar Sonar Navigation) on GNSS-R and LEO passive radar | Annotated bibliography and capability maturity assessment, updated quarterly |
| GNSS-R oceanographic baseline (CYGNSS/Spire) | Delay-Doppler map processing using published CYGNSS Level 1 data; comparison with Spire GNSS-R products for surface roughness and wind speed | Regional ocean surface wind and roughness data layer; contextual input for maritime surveillance planning |
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.