ELINT · Electronic Intelligence

Galileo’s First Authenticated Fix, and the Years Between Now and It

Under live interference in Arctic Norway, five Galileo satellites delivered the first civil authenticated position using the Signal Authentication Service. Validation begins in 2027. A British regulator's answer for the interval is radar, visual fixing and celestial navigation.

What was demonstrated

On 16 September, for about two hours, five operational Galileo satellites transmitted a new encrypted signal while receivers at Andoya in Arctic Norway and at ESA’s navigation laboratory in the Netherlands computed positions from it: “the first real-world positioning using Galileo’s upcoming SAS” [1]. EUSPA announced the milestone on 17 September as the “first-ever civil authenticated GNSS position using Signal Authentication Service (SAS)”, achieved by combining SAS with the existing navigation message authentication service [4].

The venue matters as much as the result. The test ran inside Jammertest, the annual Norwegian campaign where navigation equipment is deliberately exposed to interference under controlled conditions. The 2026 edition was held at Bleik on Andoya from 14 to 18 September [6], organised by Testnor with ESA participating since 2023 and the Norwegian communications authority coordinating [2][6]. Scenarios cover jamming, spoofing and meaconing, the rebroadcast of captured signals to produce a wrong time or position [2].

Two different things being authenticated

The distinction between the two services is the technical core of the story, and it is routinely collapsed in summaries. OSNMA, the Open Service Navigation Message Authentication, gives receivers “the assurance that the received Galileo navigation message is coming from the system itself and has not been modified”, authenticating the message broadcast on the E1-B signal component [3]. It adapts the TESLA broadcast authentication protocol, and its initial service “was declared operational on the 24th of July 2025” [3].

SAS addresses the other half. It extends authentication to the ranging component, the measurement of distance to each satellite, rather than the data the satellite sends about itself [5]. One account states the reason crisply: “verifying a message does not, by itself, verify every measurement used to calculate a position” [5]. A spoofer that replays or synthesises signal timing can leave the message content untouched and still move the receiver.

Hence the framing of the combination by ESA’s Jorg Hahn: “By combining signal authentication and navigation message authentication, users will know they can trust the position their device is showing” [1]. The mechanism described publicly is a delayed check: a receiver records a short sample of the encrypted transmission and later compares it against an expected sequence once the authentication information becomes available [5]. Key management detail is not in the public material.

Authentication is detection, not availability

This is the point most likely to be lost between the announcement and a board paper. Signal authentication tells a receiver whether to trust what it is being given. It does nothing to keep the signal there. One account states the limit explicitly: authentication “cannot guarantee that usable transmissions remain available” [5].

That is not an argument against it. The same account makes the complementary case: “Detecting that measurements are untrustworthy can itself be useful even when the receiver cannot produce a replacement solution” [5]. A bridge team told that its position is unverifiable is better placed than one steering confidently to a fabricated fix. But authentication converts a silent integrity failure into a declared outage, and an organisation still needs an answer for the outage.

What a receiver has to do to use it

Neither service is a firmware flag. The service centre documentation sets out what OSNMA asks of equipment: receivers must “Synchronize to the Galileo System Time (GST) before receiving and processing OSNMA information”, with “the time synchronization requirement set between 30 and 300 seconds”; they require “Receiver initialisation: installing and storing the certified cryptographic material”; and they need the capability for “Receiver key renewal/update” [3].

Those are procurement and lifecycle obligations, not features. A receiver holding certified key material needs a route for renewing it, which means a maintenance process for every unit in a fleet. One analysis makes the point that “a receiver’s ability to calculate a location and its ability to authenticate that calculation should be stated separately” [5].

That is the specification line to add to tenders now. A supplier claiming Galileo support is not claiming authentication support, and a receiver that authenticates internally but passes no result to the bridge display or the timing system has solved nothing a watchkeeper can act on.

The interval, and what a regulator asks for in it

SAS is not a service yet. Validation and accreditation activities begin in 2027, run by EUSPA with the European Commission, ESA and industry, before any operational declaration; both services are to be free and worldwide once declared [1][5]. One analysis adds the caution that a planned date “should be treated as a planning assumption” [5].

Which leaves an interval of years, and at least one regulator has written down what to do in it. On 1 October the UK Maritime and Coastguard Agency published MGN 719 (M+F), “Operating in GNSS Denied and Degraded Environments”, addressed to shipowners, operators, masters, officers of the watch, harbour and port authorities, VTS providers and other maritime stakeholders reliant on PNT services [8].

The substance is unglamorous and specific: review risk assessments and identify which systems depend on positioning data; establish “procedures for detecting and responding to interference”; exercise “contingency procedures in drills”; and maintain proficiency in “radar navigation and position fixing, visual techniques, celestial navigation, dead reckoning” [7]. The guidance notes a marked increase in interference reports worldwide, notably around the Black Sea, the Baltic and the Gulf, and cites a 2023 estimate that a seven-day GNSS outage would cost the UK maritime sector about GBP 1.5 billion [7].

There is a useful asymmetry here. The cryptographic fix is three or more years from operational declaration and will then require a fleet refresh. The interim measure is training and drills, and it is available this quarter.

The threat is not standing still

Meanwhile the problem is acquiring new geometry. Work by researchers at the University of Texas at Austin and Stanford, reported in June, attributed episodes of GNSS interference recorded simultaneously across Europe, Canada and Greenland since 2019 to three satellites of a Russian high-orbit constellation, after building “a framework to detect interference based on received signal power” [10]. The incidents were brief and “almost always during business hours (UTC timezone)”; the authors described space-based interference as “a qualitative escalation in GNSS interference” with “vast geographic reach”, and did not speculate on intent [10].

Terrestrial jamming is bounded by the horizon. An emitter in a high orbit is not, which is precisely the case authentication answers better than geography does — and precisely the case that cannot be mitigated by moving the route.

What to do

  • Separate the two questions in every specification: can the receiver compute a position, and can it authenticate one [5].
  • Ask suppliers which display, alarm or log an authentication result actually reaches, not merely whether the receiver supports the service [5].
  • Plan for key material: certified cryptographic material needs a renewal process for every unit in the fleet [3].
  • Treat authentication as converting silent error into declared outage, and write the response to the declared outage [5].
  • Act on MGN 719 now: risk assessment, interference detection and response procedures, drills, and crew proficiency in radar, visual, celestial and dead-reckoning navigation [7][8].
  • Report interference through the mechanism that exists; three UN agencies asked states to implement reporting and to keep conventional navigation infrastructure available [9].

The milestone is real and worth marking: for the first time a civil user computed a position from Galileo signals it could verify, while interference was being radiated nearby on purpose. The honest reading is that this is the beginning of a procurement cycle rather than the end of a problem. Everything useful between now and an operational service is the unfashionable part — specifications written carefully, drills actually run, and watchkeepers who can still fix a position without a satellite.

Sources

Every R3KONX article cites its primary material. 10 sources, in order of first citation. Links open the original publication.

  1. Galileo achieves first civil authenticated position fix under spoofing conditions GPS World · 2026-09-21
  2. Jammertest 2026 provides unique PNT testing environment GPS World · 2026-10-05
  3. Galileo Open Service Navigation Message Authentication (OSNMA) European GNSS Service Centre · 2026-10-07
  4. Galileo achieves first-ever civil authenticated GNSS position using Signal Authentication Service (SAS) EUSPA · 2026-09-17
  5. What Does Galileo Signal Authentication Mean for Trusted Satellite Navigation? New Space Economy · 2026-09-17
  6. What Does Jammertest Reveal About Europe's Satellite Navigation Resilience? New Space Economy · 2026-10-03
  7. MCA Issues Guidance on Operating in GNSS-Denied and Degraded Environments Inside GNSS · 2026-10-02
  8. MGN 719 (M+F): Operating in GNSS Denied and Degraded Environments Maritime and Coastguard Agency, GOV.UK · 2026-10-01
  9. Protect satellite navigation from interference, UN agencies urge ICAO, with ITU and IMO · 2025-03-25
  10. GPS jamming traced to Russian satellites in new 'space interference' concern Computing · 2026-06-08

Researched and written by the R3KONX analysis desk from the cited primary material. The demonstration is reported as the agencies and the trade press described it; ESA's own image caption could not be retrieved from this desk, so EUSPA's notice and two independent trade accounts are used instead. Service dates are planning assumptions stated as such by the parties involved, not commitments. The MGN 719 text is cited from its publication record and from reporting that quotes it, the full guidance PDF having returned a not-found error. 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.