Unraveling a continent-wide GPS interference mystery: Kosmos 2546 and the hunt for a space-based signal jammer

Overview

Veritasium examines a surprising GPS interference event that affected receivers across Europe. Researchers Todd Humphries and Zach Clements analyze years of GNSS data, uncovering abrupt drops in signal to noise ratio that point to a space-based source rather than solar activity or local ground-based jammers. The investigation culminates in identifying a Russian satellite, Kosmos 2546, as the most plausible culprit, while leaving room for questions about whether the signals were tests or covert communications. The video also explains how GNSS works and why resilience is essential for modern infrastructure.

• Continent-wide GPS disruptions traced to a high-altitude source, not a local one
• Time-difference measurements tie the interference to Kosmos 2546
• Discussion of future-proofing navigation with terrestrial and alternative timekeeping methods

Introduction and context

The video begins by laying out how GPS and GNSS systems provide timing and positioning services widely used in aviation, shipping, finance, telecoms, and everyday apps. Veritasium introduces the central mystery: on two specific days and times in 2021 data, receivers across a large network showed synchronized drops in signal to noise ratio, suggesting interference that overwhelmed navigation signals. The question is whether this interference came from the ground, from the Sun, or from space, and what the source would imply for global positioning reliability.

How GPS works and why it’s vulnerable

A concise primer shows that receivers determine position by measuring the time-of-flight from satellites whose signals carry precise timing. Solving for position and clock error typically requires at least four satellites, with additional satellites providing better accuracy. The signals are incredibly weak when they reach Earth, making them susceptible to jamming if someone transmits in the same frequency band. The video stresses that modern GNSS uses a protected, narrow frequency band and that spoofing or jamming can have widespread consequences for many services if exploited at scale.

The clues from the data

Humphries and his student Zach Clements first suspect ground-based interference might be responsible, given the Kaliningrad region’s tactical relevance. However, they realize the observed pattern is continental in extent and simultaneous across Europe, which is difficult to reconcile with a local ground transmitter. The geometry implies a source high above the Earth, capable of reaching receivers across long distances without being blocked by the curvature of the planet. Conservative geometry estimates place the source well above the horizon for all affected stations, at least 1200 kilometers up, well beyond the altitude of typical manned orbits or the ISS.

Space-based interference vs solar interference

The researchers rule out solar radio bursts because the bursts are abrupt and narrow in bandwidth, centered at a GPS band. Solar events typically affect large sunlit swaths of Earth and span a broad range of frequencies, which is inconsistent with the observed narrow 5 MHz slice. The evidence pushes toward a satellite-based interference source rather than solar activity or a purely terrestrial transmitter.

The narrowing search and the breakthrough

Initial analysis narrows the field to satellites that could simultaneously reach all stations, leaving about 14 possibilities, including a Russian satellite network and a Algerian satellite that initially seemed plausible. Yet, when the team examines data from GNSS receivers tracking these candidates, the Algerian satellite’s signals look like ordinary GPS transmissions that were also jammed, implying it was a victim rather than the source. Over four months, the team confronts the challenge of relying on one satellite as the interference source, an assumption that could bias the search. They pivot to a new approach: capture the raw radio signal with high temporal resolution so they can measure the precise arrival times of the jam signal at different receivers, not just the strength of the GPS signals measured once per second by standard receivers.

The critical observation and the arrival-time technique

In September 2025, at a navigation conference in Baltimore, the team unveils the plan to deploy specialized receivers across Europe to record raw waveforms at tens of megahertz. Then, in February 2026 two stations Amsterdam and Trondheim capture a 2.3 second interference event with high temporal fidelity. By aligning the raw data streams, they measure the tiny differences in arrival time between the two stations. For example, the jamming signal arrived at Trondheim about 139 microseconds before Amsterdam. This timing difference constrains the source location to lie on a hyperboloid surface in space, and the source must stay aligned with this surface as the satellite moves. The precision of tens of microseconds translates into a spatial constraint of roughly five meters, a remarkable level of accuracy for a satellite-based interference source.

Testing satellites against the timing data

With the arrival-time data in hand, the researchers systematically test each satellite's known orbit. They compute the expected time difference between Amsterdam and Trondheim if the interference came from that satellite, then compare with the observed data. The test is made more stringent by using a continuous two and a half second recording, so the hyperboloid moves with the satellite and the source must remain on the surface for the entire interval. Only one satellite matches: Kosmos 2546, a Russian satellite that appears to lie on the hyperboloid across the entire time window, with agreement within about 200 meters given orbital data uncertainties. While this is compelling, the researchers caution that the analysis is recent and not yet peer-reviewed.

Context and caveats

Kosmos 2546 is reportedly part of a six-satellite constellation associated with Russia’s early missile-warning system, often described as a Molniya-like configuration that provides long dwell times over high latitudes. This constellation allows a single satellite to image large portions of the Earth, enabling wide-area jamming capabilities. However, Kosmos 2546 was launched in 2020, which complicates explanations for disruptions dating back to 2019. The team clarifies this discrepancy and notes that some other satellites in the same constellation could still be involved, or the observed bursts could be short communications from those satellites rather than continuous jamming. Independent European teams have verified certain aspects of the work, but the study remains preliminary and not yet peer-reviewed. A second interference burst at a lower frequency around 1558.5 MHz, which overlaps with Beidou signals, adds another layer of complexity and raises questions about broader, potentially coordinated space-based interference activity.

Implications and potential explanations

The investigators present two plausible interpretations. One is that the European interference bursts were tests of a space-based jamming capability, performed briefly and deliberately to calibrate the system. The other is that the bursts could represent very short, highly specific communications from satellites in Russia’s missile-warning network that incidentally affected GNSS receivers. The video emphasizes that even if the jamming is not always deployed, the mere existence of a space-based, high-power, near-Earth jammer represents a significant escalation in electronic warfare. The researchers also discuss the broader picture of GNSS vulnerabilities, noting that many navigation and timing-dependent systems would be affected if such a capability were fully deployed against civilian infrastructure.

Beyond jamming: spoofing and resilience

Veritasium transitions to discuss GPS spoofing, which replaces real signals with counterfeit ones, and notes that spoofing and jamming have different risk profiles. The video highlights the need for resilience through a multi-layered PNT architecture that does not rely solely on space-based signals. Examples from other countries and groups are mentioned, including fiber-based time transfer and terrestrial navigation alternatives. The conclusion stresses the importance of preparing for a future in which navigation and timekeeping could be disrupted by space-based capabilities, and outlines potential paths forward such as ground-based timing, optical fiber networks, and magnetic or quantum-based positioning concepts. The piece closes with a reflection on the central role of GPS in modern life and the potential consequences of a major vulnerability being exposed along with a path toward more robust, diversified time and navigation systems.