Space-Based GNSS Jamming Exposes DP Integrity Monitoring Blind Spots

What Happened

Researchers at the University of Texas Radio Navigation Laboratory have identified a constellation of Russian early-warning satellites in Molniya (“lightning”) orbits as the source of difficult-to-detect disruptions to GPS signals. The disruptions have been recorded since at least October 2019. Zachary Clements, Argyris Kriezis, and Professor Todd Humphreys describe the phenomenon in a paper titled “Chasing Lightning,” documenting 75 events over a roughly seven-year period that affected reference stations in Europe, Greenland, and Canada.

The team analysed data from 165 reference stations of the International GNSS Service. In each event they observed a synchronous drop in carrier-to-noise ratio – in some cases by as much as 10 dB – at stations thousands of kilometres apart. The events are short-lived, typically under 10 seconds, and tend to fall on weekdays, most often Tuesday through Thursday. According to the published analysis, the interference is centred near 1577.5 MHz, roughly 2 MHz above the GPS L1 carrier, degrading L1 while leaving the L5 band unaffected. Humphreys has characterised the pattern as too consistent to be accidental, stating the data indicates it has to be intentional.

The work was prompted by observations presented at a September 2024 Civil GPS Service Interface Committee meeting in Baltimore, with the team’s findings later presented at an Institute of Navigation conference in 2025 before the full paper appeared in 2026. The same constellation has been reported to affect China’s BeiDou signals in a similar way since June 2020.

Why This Matters for Offshore Positioning

DP integrity rests on a single principle: multiple independent position reference systems are voted against each other. If several GNSS receivers agree within tolerance, the DP control system treats the solution as healthy. If one receiver disagrees, it is flagged and weighted out. Reference systems are expected to fail individually – a lost hydroacoustic transponder, a drifting inertial unit, a loss of differential correction. These are local problems, and the voting model is designed around them.

A space-based source that degrades many GNSS receivers simultaneously over a wide area undermines the independence assumption the voting model rests on. Consider a DP2 intervention vessel holding position 500 metres from a subsea structure with four GNSS-based reference systems. During a short L1 interference pulse, all four receivers degrade together. If accuracy worsens across every receiver at once – say from sub-metre to a few metres – the receivers still agree with one another, so the voting logic sees consensus rather than fault. The vessel keeps working while positional uncertainty is materially larger than the control system reports. A 10-second event is often too brief to trip a position-loss alarm, and the residual shows up in logs as noise or multipath rather than as a systematic loss of accuracy. The DP reference filter smooths a single brief excursion, so the modelled position barely moves; the concern is not one isolated pulse but repeated transients, or one that lands during a close-quarters phase, when even a few metres of unflagged uncertainty eats directly into the working margin.

Why Single-Failure Analysis Misses This

Worst-Case Failure Design Intent (WCFDI) is the basis of Failure Mode and Effects Analysis (FMEA) for DP systems. The idea is to identify single failures and provide redundancy. A switchboard fails – there is another. A thruster trips – there are others. The position reference is degraded – different reference-system types provide diversity.

Space-based interference targets specific GNSS frequencies and creates a failure mode that single-failure analysis does not model. IMO guidelines for DP vessels, currently MSC.1/Circ.1580, are framed around position keeping after a single failure. What they do not address is an external factor that degrades an entire class of reference systems at once. Conventional DP integrity checks monitor sensor disagreement; they are not designed to detect a widespread, modest degradation that affects every GNSS input in step. The system assumes one receiver is at fault when readings diverge; it does not assume all readings can drift slightly together. That is the gap. During critical operations – deploying a submersible near subsea structures, holding position over a wellhead, or maintaining station during diving in the 500-metre zone – a transient error of two or three metres may not cause an incident, but it erodes the safety margin those operations depend on.

Where Operators Get It Wrong

1. Assuming More GNSS Receivers Equals Better Integrity

Many project specifications call for four or six GNSS receivers in DP2/DP3 mode. The logic is that more receivers mean more tolerance to failure. That holds for independent failures – a receiver hardware fault, a damaged antenna cable, a lightning strike on an antenna. It does not hold when interference affects a shared frequency. If an L1-band source degrades the L1 signal universally, six receivers operating only on L1 are no more resilient than two. Different manufacturers, different firmware, different signal processing – all of it shares the same vulnerable L1 input.

2. Treating Multi-GNSS as Full Independence

Position reference systems often use GPS, GLONASS, Galileo, and BeiDou together. Risk assessments sometimes claim that multiple constellations deliver independent positioning. That is partly true for satellite geometry and for localised outages – a gap in GPS coverage does not by itself affect GLONASS – but not when an interference source targets a shared frequency band. GPS L1 and Galileo E1 are both centred at 1575.42 MHz, while GLONASS G1 sits near 1602 MHz. A source covering the L1/E1 band degrades GPS and Galileo together; constellation diversity does not protect against frequency-band interference. The independence is partial, not absolute.

3. Believing “We Would Have Seen It” in Post-Mission Data

Teams often assume that any accuracy degradation leaves a clear trace in DP logs. A brief common-mode GNSS degradation can instead surface as residual position values that are easily read as vessel motion or as multipath from crane operation. Performance plots may dip briefly, but unless you log and trend carrier-to-noise ratio across receivers – not just position – a simultaneous accuracy loss can pass unremarked. Standard DP reports focus on alarms, power events, and system faults. Sub-alarm degradations that never trip a threshold are easy to miss.

4. Assuming Frequency Diversity Solves the Problem

The reported interference affects GPS L1 but not L5. Modern multi-frequency receivers with L1/L5 support are better placed than L1-only receivers, and exploiting an unaffected band is a genuine mitigation. But it is not a cure. Not all DP reference systems use multi-frequency receivers, and not every region has mature multi-frequency correction infrastructure. Frequency diversity reduces exposure to this specific source; it does not remove the broader risk, since a capable source could target other bands.

Effects for DP Integrity Monitoring

The core issue is common-mode interference. DP integrity monitoring scans for sensor disagreement, so it can miss the case where every GNSS receiver degrades together yet remains mutually consistent. IMCA has begun to address this directly: its guidance on satellite-based positioning for offshore applications (IMCA M 242) and its information note on GNSS jamming and spoofing for DP cover GNSS vulnerabilities including interference and spoofing, and recommend multi-frequency receivers and mixed reference systems combining GNSS with hydroacoustics and radar. What remains under-developed in everyday practice is the monitoring discipline for wide-area, transient, space-based degradation specifically.

A practical step is to monitor carrier-to-noise ratio across all GNSS receivers rather than relying on position alone. A correlated CNR drop across multiple receivers is a warning sign of an external cause even when the position solution still looks consistent. That signature is different from normal environmental variation, which affects antennas unevenly, and different from a single-device fault, which is isolated to one receiver. Many DP systems log this data, but cross-receiver trend analysis is not yet common practice – CNR plots tend to be read per receiver, for receiver health, rather than across receivers for a shared external event.

What Operations Teams Should Do Now

First, confirm whether your GNSS reference receivers operate on multiple frequency bands – for example L1/L5 for GPS, E1/E5a for Galileo, and G1/G3 for GLONASS. If a unit is L1-only, treat it as exposed. Equipment predating widespread multi-frequency support should be assumed L1-only until verified otherwise.

Second, baseline and monitor carrier-to-noise ratio. Establish normal CNR values before critical work. A simultaneous, sustained CNR drop of 3 dB or more across receivers during operations warrants attention even if coordinates are not visibly moving. Brief excursions do not call for panic, but they do call for a check that positional uncertainty still fits the operation’s tolerance.

Third, take common-mode GNSS degradation into the FMEA. Most analyses treat GNSS reference loss as either a total dropout or a single-receiver fault. Simultaneous, partial degradation of all GNSS inputs should be added as a credible scenario, with non-GNSS references – hydroacoustics, taut wire, radar-based systems – available as backup. For operations with tight positional tolerances such as subsea work, well operations, or heavy lifting, verify that position control can be maintained without GNSS. If station keeping on hydroacoustics and gyrocompass alone has been demonstrated, GNSS degradation monitoring should be treated as routine practice rather than a contingency.

Finally, when planning, weigh signal independence, not just sensor count. “Six GNSS receivers” sounds robust, but not if they all share the L1 band. A configuration such as “two multi-frequency GNSS, plus an acoustic array and an independent surface reference” delivers far more genuine independence.

Standards Will Follow the Evidence

Industry standards move with evidence, and that takes time. IMCA, IOGP, and IMO will continue to refine guidance as space-based interference becomes better characterised, building on existing material such as IMCA’s satellite-positioning guidance and its jamming-and-spoofing note. In the meantime, the responsibility sits with operators to interpret today’s requirements with a clear view of how the threat landscape has widened.

This is not science fiction. A peer-scrutinised analysis has documented 75 wide-area events over roughly seven years, affecting reference stations across continental scales – transient, hard to detect, and targeted at a specific frequency band. Systems built around independent, fault-tolerant references were not designed for a source that degrades a whole class of inputs at once. The lesson is not to abandon GNSS but to respect its limits: diverse, genuinely independent references remain the right approach, and simultaneous GNSS degradation is now a documented reality rather than a hypothetical. If fault detection leans entirely on GNSS receivers checking each other, that assumption deserves a hard look now.

---

Based on: Russia Has Been Jamming GPS From Space Since 2019