Terrestrial Timing Networks: What NPL-Vodafone Means for Offshore PNT

What Happened

NPL and Vodafone have completed trials of a telecom version of NPLTime as an alternative timing source to GNSS in mobile network synchronisation. In the UK, VodafoneThree is the first mobile network operator to test a terrestrial time source from the national metrology institute alongside GNSS in its network timing infrastructure. NPLTime is an end-to-end, fibre-based timing service traceable to UTC(NPL); NPL has run the UK’s national time scale for around 30 years and has disseminated NPLTime to the finance sector for regulatory compliance for roughly eight years. The telecom version is designed to meet International Telecommunication Union (ITU) standards for accuracy, stability, resilience and traceability, and to maintain accuracy within 40 nanoseconds.

The stated aim is to remove GNSS as a single point of failure for positioning, navigation and timing (PNT) services that underpin critical infrastructure. Vodafone says it intends to reduce GNSS reliance across its European markets and to replicate the same timing infrastructure beyond the UK as 5G matures and 6G is planned.

Why This Matters

The maritime positioning chain is fragile in a familiar way: when GNSS degrades, much of the supporting infrastructure degrades with it. The NPL-Vodafone trials show that ground-based synchronisation over fixed fibre can stand in for GNSS-derived time onshore. You will not run fibre to a drillship 200 kilometres offshore. But the land-based systems that generate PPP corrections, broadcast SBAS, and hold the CORS reference network together are themselves dependent on accurate timing, and where that timing is itself derived from GNSS it inherits the same vulnerability. If a shore reference loses its timing, whether to a wide-area disturbance or to interference at that site, and there is no terrestrial backup, the position offshore is at risk no matter how many receivers the vessel carries.

A typical offshore positioning chain runs: GNSS receiver on board to land-based correction service, correction data transmitted back to the vessel, processing through a Kalman filter, final position. The correction side demands very high timing accuracy to model the ionosphere and correct satellite orbits. When telecom operators harden their synchronisation, the positioning services that lean on the same reference infrastructure stand to benefit. That linkage is the reason a telecom timing story is also an offshore PNT story.

The Offshore PNT Dependency Problem

In offshore operations, GNSS is too often treated as infallible. Two GNSS receivers on a DP vessel are spoken of as redundant; a few more on a survey vessel and the failure modes are assumed covered. But when GNSS is jammed or spoofed, every one of those receivers can show the same corrupted signal. They are not independent sources. They share the same satellites, the same ionosphere, and the same interference.

GNSS is weak under contested or disturbed conditions. L-band signals are low-power and easily overwhelmed; jamming and spoofing are now achievable with commercial radio equipment. The 2017 Black Sea spoofing incident near Novorossiysk reportedly caused around twenty vessels to register false positions onshore, illustrating how a single spoofing source can mislead multiple receivers at once. No amount of receiver redundancy protects against that, because the threat is to the signal, not the hardware.

Yet contracts routinely demand high survey accuracy while leaving GNSS as the primary, unbacked position sensor. IHO S-44 Order 1a sets a total horizontal uncertainty (THU) of 5 m plus 5% of depth, but that is the most relaxed of the orders likely to appear in an offshore spec. Many infrastructure and engineering surveys invoke Special Order (THU 2 m) or Exclusive Order (THU 1 m), where the positioning budget is far tighter and GNSS-timing dependence bites correspondingly harder. The point is not the headline number but that the entire uncertainty budget rests on a single position sensor. Jamming, spoofing, and ionospheric disturbance from solar activity can all degrade GNSS in otherwise benign conditions. Terrestrial timing does not solve positioning on the open sea, but it sharpens the question of where the single points of failure really sit.

Where Operators Get It Wrong About Positioning Resilience

1. Confusing Receiver Redundancy with Signal Redundancy

A DP FMEA that counts GNSS receivers is not measuring source redundancy. Multiple receivers tracking the same constellations, through the same ionosphere, exposed to the same interference, are one source, not several. True redundancy requires different physical principles: acoustic positioning, inertial navigation, or radio positioning on different frequencies. LBL arrays provide this if you are prepared to deploy on the seabed. Inertial systems drift without external aiding. eLoran operates at 100 kHz and is immune to L-band interference, but coverage is patchy and integration is non-trivial. Class rules treat GNSS as one position reference type for that reason: a DP2 station-keeping case generally needs independent reference types, not just more GNSS receivers.

2. Assuming PPP and RTK Corrections Will Always Be Available

PPP and RTK corrections depend on coastal reference networks, and those networks depend on reliable timing. Two distinct failure points are easily conflated. The first is a GNSS outage at the rig itself, where regional jamming or spoofing corrupts the vessel’s own receivers while the shore reference station, hundreds of kilometres away, still tracks clean multi-band signals and derives valid GNSS time. The second is a loss of timing at the shore reference station, whether from a wide-area space-weather event or interference local to that site, which degrades the correction stream even when the vessel’s receivers are healthy. The two are independent, and a resilience case has to address both rather than assume one covers the other. Operators tend to assume commercial services such as Fugro Marinestar and Veripos handle the second case seamlessly. Those providers do run multi-constellation networks and ground infrastructure, but they rarely publish detail on their ground synchronisation, so the degree to which their reference timing would survive a GNSS event is a posited vulnerability, not a demonstrated one. It should be treated as a question to put to the provider, not an established failure mode. Fibre timing of the kind NPL and Vodafone are trialling is the kind of terrestrial reference a correction provider could in principle build on, and if it did, offshore users would benefit at one remove.

3. Treating Timing Accuracy as a Non-Issue for Positioning

PNT performance depends directly on timing, because position is derived from signal propagation time. GNSS receivers need timing accurate to fractions of a microsecond. Survey-grade time stamping is tighter still, and multibeam echosounders need precise time alignment with motion sensors. When a specification calls for “GNSS with external 1PPS synchronisation,” GNSS quietly becomes the master reference for both position and time. As GNSS timing degrades, positional accuracy degrades with it. Terrestrial timing can decouple these dependencies onshore. Offshore, that decoupling is much harder to achieve.

4. Ignoring What Happens to Shore Infrastructure During GNSS Outages

DGPS, AIS, and VHF base stations all rely on GNSS timing. Solar disturbances and deliberate interference disrupt GNSS, and without a terrestrial timing backup the shore stations become as unreliable as the vessels they serve. The NPL-Vodafone work addresses the telecom sector; maritime shore infrastructure has no equivalent programme. As an illustrative mechanism rather than a documented case, consider a vessel in the North Sea using DGPS from a base station 50 km away. If a space-weather event disrupts GNSS and the base station loses timing, correction latency rises, message integrity suffers, and the vessel may see unstable-DGPS warnings without immediately knowing the cause until it cross-checks against an independent reference. Terrestrial timing cannot restore the vessel’s own GNSS, but it can keep correction services stable.

The eLoran Question

eLoran is the only terrestrial timing and positioning technology with genuine offshore reach. It operates at 100 kHz from coastal transmitter towers and covers hundreds of kilometres offshore. Timing accuracy is roughly 50 to 100 nanoseconds, and positional accuracy in coastal waters is about 10 to 20 metres, adequate for DP station-keeping, survey line approach, and transit, but not for end-product survey positioning.

The picture is uneven. The UK’s Loran-C service was switched off at the end of 2015. The United States retains transmitter sites that could support eLoran but they are not in operational service. South Korea operates an active eLoran network. ITU standards exist, but deployment lags. If governments choose to back terrestrial PNT for long-term resilience, eLoran offers the best route to offshore coverage. For now, the momentum is in terrestrial fibre timing, which is well suited to fixed land installations and of no direct use to a vessel at sea.

Practical Positioning Resilience Offshore

Terrestrial timing infrastructure will not be deployed on offshore platforms in any near term. The realistic levers are about hardening what exists.

• Use multi-constellation GNSS receivers. GPS, GLONASS, Galileo, and BeiDou operate independently, so degradation of one constellation need not cause a position failure. New vessels generally have this; many older vessels still rely on GPS alone.
• Keep acoustic positioning available as a backup, calibrated while GNSS is healthy. LBL or USBL arrays provide independent positioning from acoustic time-of-flight, slower, less accurate and more expensive than GNSS, but genuinely independent. Detecting acoustic errors after GNSS has already failed is far harder than verifying the array beforehand.
• Specify shore correction-service resilience in contracts. Ask your PPP or RTK provider how they handle GNSS outages: do they hold a terrestrial time reference, do they validate corrections during space-weather events, and what is the backup accuracy? Define what an acceptable answer looks like rather than just collecting the questions. A workable benchmark is that the reference timing must hold within the order of 100 nanoseconds for a defined hold-up period through a GNSS outage, with stated, bounded degradation of the correction accuracy over that window rather than an open-ended fail. Treat “the corrections simply stop” or “we cannot quantify the degradation” as a failing answer, and require the provider to state the assumptions behind any guarantee. Providers rarely publish this, but they will answer when asked.
• Monitor GNSS integrity, not just availability. Standard bridge equipment shows satellite count and HDOP but says little about signal quality. Survey receivers expose SNR, multipath, and ionospheric delay; watch these continuously, because spoofing and jamming move integrity metrics before position visibly drifts.
• Plan for sustained outages. What is the fallback if GNSS is down for 24 hours? Can DP performance hold on radar, laser, and microwave references alone? Can survey continue on inertial and acoustic systems? Assuming rapid GNSS recovery is the weakest part of most contingency plans.

What the NPL-Vodafone Work Means Long-Term

Telecom operators have to harden ground synchronisation because 5G and 6G cannot tolerate GNSS outages. NPL and Vodafone are demonstrating fibre-based timing links that can scale across Europe, and that creates a foundation other sectors can draw on. Maritime reference systems, DGPS base stations, and correction services could in principle integrate terrestrial timing and quietly improve offshore reliability.

The shipping industry will not drive this investment. Offshore operators are users, not infrastructure owners; their leverage is in contracts, equipment standards, and procedures. If telecom companies and metrology institutes expand resilient PNT, offshore benefits as a second-order user. The trials show that terrestrial alternatives to GNSS exist and are being deployed in fixed installations. That is real progress, but it is progress that mostly stops at the shoreline.

For offshore operations, the honest conclusion is that resilient PNT is still some way off. Fibre timing works where cables run. eLoran can span open water but needs government investment. LEO-based PNT solutions are still years from maturity. Until then, GNSS remains the primary offshore sensor with inadequate backup, and the right posture is to assume satellite timing can fail rather than to assume it never will.

---

Based on: NPL collaborates with Vodafone on terrestrial timing