GPS jamming has gone from a military nuisance to a daily hazard for commercial aviation. Over the past two years, reports of GNSS radio-frequency interference — the deliberate or accidental flooding of the L1 band that carries civilian positioning signals — have multiplied around conflict zones, near military exercise areas, and even in the vicinity of civilian airports. Pilots receive warnings, aircraft switch to inertial backup, and on bad days flights divert. The hard problem is detection at scale. Building a dedicated sensor grid to monitor interference across a continent would cost billions and take a decade. A team of researchers publishing a 2026 arXiv preprint proposes a cheaper answer: repurpose the 1090 MHz receivers that already blanket the sky, the ones feeding FlightRadar24 and OpenSky Network, and turn them into an opportunistic, continent-scale GNSS interference monitor. [arXiv:2607.09700]
The Core Finding
The paper proposes a three-stage framework. The first stage screens every flight report for position-jump candidates — the signature of a receiver being told, by a jammed or spoofed GNSS feed, that the aircraft is somewhere it is not. The second stage verifies local track consistency to suppress false alarms caused by heterogeneous receiver timestamping, a well-known problem when different brands of ADS-B boxes use different GPS pulse-per-second alignments. The third stage groups confirmed anomalies into traffic-adaptive multi-aircraft events, so a single jamming incident affecting dozens of flights becomes one cluster rather than dozens of noise points. Think of it like a fact-checker who first flags suspicious claims, then cross-references them against independent witnesses, then clusters related incidents into a single coherent story.
The headline result is concrete. Applied to 605 million ADS-B reports collected over Northeast Asia between December 2025 and February 2026, the framework identified 166 event clusters inside the validity window of Notice to Airmen RKRR Z1401/25 — a published Korean military exercise restriction — and zero clusters in the pre-NOTAM baseline period. The framework's most counterintuitive finding is that more than 99 percent of confirmed anomalies remained in high quality-indicator regimes, meaning the standard ADS-B accuracy metrics that air traffic control relies on would have missed them. Track-consistency verification, the authors argue, "provides a complementary sensing criterion for GNSS RFI monitoring" — a sentence that quietly indicts a decade of relying on NACp values alone.
The State of the Field
Existing GNSS interference monitoring leans on dedicated ground stations, high-rate GNSS reference networks, or satellite-borne payloads. These are accurate but sparse. A typical national RFI monitoring grid might have a few dozen sensors; a crowdsourced ADS-B network has tens of thousands, and they are already installed, already streaming, already paid for. The trick is that ADS-B was designed for surveillance, not for physics. Its Navigation Accuracy Category for Position flag, NACp, tells you how confident the avionics are in the reported fix, but the avionics have no way to know that the satellites above them are being jammed. A confident receiver reporting a confident position over a military exercise area looks identical, in NACp terms, to a confident receiver reporting a confident position over a quiet airport. Prior crowdsourced RFI work has used simple position-jump thresholds, which work in clean data and fall apart when receiver clocks disagree by a few milliseconds. The 2026 paper's contribution is the three-stage pipeline that adds physical-plausibility checks and multi-flight clustering on top of the raw jumps.
The timing matters. GNSS interference incidents have increased sharply since 2022, particularly over the Black Sea, the Baltic, the Eastern Mediterranean, and the Korean Peninsula. Airlines, regulators, and military planners all want the same thing: a live map of where GNSS is being denied or spoofed, updated faster than the next NOTAM cycle. Crowdsourced ADS-B is the only data source dense enough to deliver that, if the noise can be filtered out. This paper shows the filtering works.
From Lab to Reality
For aviation regulators, the immediate path is integration with NOTAM workflows. The framework's ability to confirm that 166 event clusters lined up with a single published military exercise, and that zero clusters appeared before the NOTAM went live, is exactly the kind of validation an agency like the FAA, Eurocontrol, or Korea's KAC would want before signing off on an operational tool. Air navigation service providers could fold crowdsourced RFI monitoring into existing surveillance dashboards, painting interference heatmaps over their flight information regions in near real time. For airlines, the practical benefit is pre-flight risk assessment: a route through a region currently flagged as having active interference can be re-planned or crew-briefed before pushback.
The commercial market is broader than aviation. Autonomous trucking, precision agriculture, maritime pilotage, and military logistics all depend on GNSS. The market for GNSS anti-jamming hardware and interference-monitoring services is already measured in the hundreds of millions of dollars annually and growing double-digit year over year. A software layer that turns existing crowdsourced infrastructure into a monitoring product is a much smaller investment than a new satellite constellation, and it could reach production inside two years if a major receiver network or aviation authority adopts it.
What Still Needs to Happen
Three technical gaps remain. First, the method is bounded by receiver density. Over the open Pacific, the polar routes, sub-Saharan Africa, and most of the Southern Ocean, ADS-B coverage is too thin to detect interference. The OpenSky Network and FlightRadar24 continue to add ground stations, but until satellite-based ADS-B reception matures — projects like the Aireon space-based payload help, but they cost real money — oceanic gaps will persist. Second, the 2026 study validates the framework against one known NOTAM event in one region. Generalizing to spoofing attacks, to wideband versus narrowband jammers, and to other geographies requires more labeled ground truth, and groups such as the German Aerospace Center's Institute of Communications and Navigation and Stanford's GPS Laboratory have both published on complementary detection methods that could be merged with track-consistency analysis. Third, the 99 percent finding is operationally awkward: it implies that the quality metrics currently trusted by air traffic control are systematically underreporting interference. Changing ATC procedures is measured in years, not months, and a full operational deployment is realistically three to five years away, with intermediate steps in advisory and post-flight analysis roles sooner.
Conclusion
In short: this 2026 arXiv paper shows that crowdsourced ADS-B receiver networks, processed through a three-stage track-consistency framework, can detect and characterize GNSS radio-frequency interference at continental scale — 605 million reports, 166 confirmed event clusters, and a clean zero in the pre-NOTAM baseline.
Frequently Asked Questions
What is GNSS RFI?
GNSS radio-frequency interference is any signal that disrupts the Global Navigation Satellite System — the family of constellations including GPS, GLONASS, Galileo, and BeiDou — that aircraft, ships, and millions of other systems rely on for positioning and timing. It can be intentional jamming, deliberate spoofing that broadcasts fake satellite signals, or accidental emissions from nearby electronics.
How does the three-stage crowdsourced framework work?
Stage one scans every ADS-B report for sudden position jumps, which are the most common signature of a flight receiving bad GNSS data. Stage two checks whether nearby reports from the same aircraft form a physically plausible track, filtering out false jumps caused by receiver clock errors. Stage three groups confirmed anomalies from many aircraft into single interference events using traffic density to set cluster boundaries.
How does this compare to traditional GNSS interference monitoring?
Traditional monitoring uses dedicated GNSS reference receivers and high-accuracy geodetic equipment. It is more precise per sensor but orders of magnitude sparser in coverage. The crowdsourced ADS-B approach trades single-point precision for continent-wide coverage, and the 2026 paper argues the two are complementary rather than competing.
When could this be commercially relevant?
Software integration into existing receiver networks and aviation dashboards could happen within two years. Operational use by air traffic control, where procedures are tightly regulated, is realistically three to five years out, beginning with advisory roles and post-flight analysis.
Which industries benefit most?
Commercial aviation gains the most direct safety benefit. Secondary beneficiaries include maritime operators in interference-prone waterways, autonomous vehicle fleets, precision agriculture, and military logistics planners who need to know when GNSS is unreliable in a given region.
What are the current limitations of this research?
Coverage is limited to regions with dense ADS-B receiver networks, the 2026 study validates the method against a single known NOTAM event, and more than 99 percent of confirmed anomalies carried high quality-indicator flags, which means current ADS-B quality metrics systematically underreport interference and would need to be re-evaluated.