ELINT · Electronic Intelligence

Commercial RF collection turns on spacecraft, and the evidence is a whitepaper

A commercial operator says its radio-frequency sensing satellites detected and characterised transmissions from spacecraft in low, medium and geostationary orbit under a US Air Force Research Laboratory contract. The claim rests on the company's own publication. No independent performance data has been released.

What was published

HawkEye 360 says its commercial radio-frequency sensing satellites detected, associated and characterised transmissions from other spacecraft in low, medium and geostationary Earth orbit. The work was carried out under an Air Force Research Laboratory contract with United States Space Force participation [1][2][5]. Payload puts the award at approximately USD 1.2 million, a Phase II Small Business Innovation Research contract originating in 2024 [2]. SatNews describes the demonstration as completing the AFRL effort, and says the company now intends to integrate space-target processing across its operational clusters [1].

The company’s 2024 announcement of that Phase II award, made through AFWERX, discloses no contract value. It describes the scope only as the delivery of its RFGeo and RFIQ data products, and quotes chief growth officer Alex Fox calling the award a milestone [6]. The USD 1.2 million figure is therefore traceable to trade reporting, not to the vendor or the awarding office.

The dates do not agree either. The company’s public executive summary is dated 29 September 2026 [3][5]. Payload and SatNews reported the release on 6 October [1][2]. KeepTrack’s daily brief dates the whitepaper release to 7 October [4]. The spread is small, but there is one publication with at least three dates attached to it.

What the system is

HawkEye 360 flies microsatellites in clusters of three. Synchronised timing allows the cluster to compare a signal’s time of arrival and frequency of arrival across three receivers and multilaterate the transmitter’s position, delivered with a 95 per cent confidence ellipse [8]. The software-defined radio tunes from 70 MHz to 6 GHz with up to 56 MHz of instantaneous bandwidth; a low-noise block downconverter extends coverage to roughly 18 GHz, so the system spans VHF through Ku-band [8].

Constellation size is unsettled in the public record. The same reference entry describes 30 microsatellites in its summary and lists 14 clusters, about 42 spacecraft, in its launch table, with launches running from December 2018 to March 2026 [8]. Operators planning around coverage should treat the number of clusters actually on station as a question for the vendor rather than a published figure.

Those sensors were built to look down at terrestrial emitters. The demonstration pointed them at orbital targets. For geostationary and medium Earth orbit the company changed its pointing technique; for low Earth orbit it developed new concepts of operation to deal with high relative velocity and short collection windows [2][4]. Chief operating officer Todd Probert told Payload that a space-based RF sensor can access signals and geometries a terrestrial sensor may not, and that some emissions are hard or impossible to observe from the ground [2][4].

What the demonstration establishes, and what it does not

The useful part of the public material is the discipline it applies to its own claim. Processing is described in three distinct stages: detection, meaning a signal was observed; geolocation, meaning an origin was estimated; and characterisation, meaning an emitter type and activity were assessed [3][5]. Those are separate problems with separate failure modes, and collapsing them is how RF claims get overstated.

Within the conditions tested, a transmission can be detected, located and characterised. What has not been shown is more consequential for a buyer:

  • That every satellite can be monitored continuously, or that a validated operational service exists [3][5].
  • That the system reads satellite traffic. Identifying an emission is not decrypting it, and nothing in the published material suggests a communications intercept capability [5].
  • Detection probability for orbits, signal types or antenna orientations outside those tested. Performance should be assessed per orbital regime, not in aggregate [3].
  • That a missing detection means anything about the spacecraft. Receiver coverage, antenna geometry and frequency sensitivity all produce the same null result as a silent satellite [3][5].

The association step is the one to watch. Several objects may fit an incomplete observation, and overlapping transmissions can confuse attribution [3]. An incorrect association does not look like an error. It looks like a satellite behaving unexpectedly. In a space domain awareness pipeline intended to run with limited human involvement [2], that failure mode propagates as a finding rather than a fault.

The collector population is commercial, and it is growing

This is not one company’s experiment. Space-based RF geolocation has become a commercial product line with multiple vendors and more than one architecture.

StarRF, founded in 2023, operates three Diting-01 satellites built by Chang Guang Satellite Technology and launched from Jiuquan in September 2026 on a Lijian-1 flight that carried nine spacecraft in total. The three work as a cluster, comparing arrival times to estimate an emitter’s position, and the company’s stated plan is a 60-satellite low Earth orbit constellation selling spectrum data to defence, public security, radio management, shipping, rail and emergency response customers. Technical details have not been disclosed, and the reported arrival-time method rests on Chinese domestic media rather than an official statement [11].

Unseenlabs reports 23 satellites launched as of 2026, adding three to six a year, with revisit of about three hours falling below one hour at full constellation, coverage of up to 300,000 square kilometres per pass, and geolocation accuracy of up to one kilometre from a single spacecraft rather than a three-satellite cluster. Its stated applications are maritime: illegal fishing, intrusion into protected waters, offshore assets, sea routes and undersea cables [10]. Spire has separately demonstrated single-satellite geolocation of S-band and X-band emitters, bands it describes as common to radar and sensing systems, on a spacecraft launched in late 2025 and co-funded through a European Space Agency contract under Luxembourg’s LuxIMPULSE programme [9].

For the Asia-Pacific the maritime application is the near-term one. Single-kilometre geolocation of ship-borne emitters across regional waters is a surveillance capability available for purchase, and it does not depend on the vessel cooperating. The orbital application announced this month extends the same logic upward.

What to do about it

For satellite and ground-segment operators, the planning assumption should change. Downlink activity, the pairing between a spacecraft and a particular ground station, and deviation from an established emission pattern are now collectable by commercial means [2][4]. Manoeuvre and payload activity that correlate with transmission changes are inferable from the outside. Emission control is no longer only a military discipline; it is an operational security consideration for any operator whose schedule reveals something it would rather not.

For buyers of space domain awareness RF services, the procurement questions follow directly from the gaps above. Ask for per-regime detection performance rather than a single figure. Ask how ambiguous associations are handled and how they are flagged downstream. Ask for observation-to-delivery latency, and for independent validation against known transmitters [3]. A capability demonstration is a starting position in a negotiation, not a service level.

For analysts, the correct reading of this month’s announcement is narrow and still significant. Radar, optical and laser ranging establish where an object is; RF establishes what it is transmitting [3]. That is a genuine addition to the catalogue, and it has arrived through a small research contract to a commercial operator rather than a programme of record. AFRL’s earlier three-year, USD 15.5 million experimental purpose agreement with the company, awarded in January 2022 for hybrid space intelligence, surveillance and reconnaissance work, is the same procurement pattern at a different scale [7].

The electronic intelligence point is the one that will outlast the press cycle. Collection against spacecraft emissions has moved from a sovereign capability to a purchasable one, and the first public evidence of it is a vendor whitepaper with no independent metrics. Both halves of that sentence should inform how it is used.

Sources

Every R3KONX article cites its primary material. 11 sources, in order of first citation. Links open the original publication.

  1. HawkEye 360 Adapts Commercial Satellite RF Sensors for Space Domain Awareness Under AFRL Contract SatNews · 2026-10-06
  2. HawkEye 360 Turns its RF-Sensors to Orbital Targets Payload · 2026-10-06
  3. What Can HawkEye 360's Radio Observations Reveal About Other Satellites? New Space Economy · 2026-10-07
  4. HawkEye 360 RF Sats Detect Other Satellites, Space Brief 7 Oct 2026 KeepTrack · 2026-10-07
  5. HawkEye 360 Demonstrates Space-Based RF Tracking of Other Satellites Israel Defense · 2026-10-12
  6. HawkEye 360 Awarded SBIR Phase II Contract to Enhance National Defense HawkEye 360 via PR Newswire · 2024-07-18
  7. HawkEye 360 Awarded $15.5 Million Contract with Air Force Research Laboratory HawkEye 360 via PR Newswire · 2022-01-11
  8. HawkEye 360 mission reference entry eoPortal · accessed 2026-10-17
  9. Spire Global Demonstrates Single-Satellite RF Geolocation Capability Spire Global · page last modified 2026-03-24
  10. Unseenlabs mission reference entry ESA Earth Online · accessed 2026-10-17
  11. China launches first satellites for 60-spacecraft radio network Interesting Engineering · 2026-09-23

Researched and written by the R3KONX analysis desk from the cited primary material. Methodological caveat: the central technical claim in this article originates in a vendor publication and has not been independently validated; no detection rates, false-positive rates or per-orbit performance figures have been published, and the contract value is traceable to trade reporting rather than to the vendor or the awarding office. Corrections to event@r3konx.asia.

R3KONX 2027

Work in this domain? So does the programme

4–6 May 2027, World Trade Centre Kuala Lumpur. Curated talks, hands-on training and The BattleGrid.