University of Texas researchers turned years of unexplained GPS interference into the first reproducible case for attributing GNSS disruption to a specific orbital asset.
For years, short, sharp interference events kept scrubbing GPS signals across a wide arc of Europe, Greenland, and the Mediterranean. Nobody could explain them. A University of Texas team now says it can, and the method matters as much as the answer.
Working from at least 75 documented 10-second high-power bursts at 1558.5 MHz — the frequency used by GPS and European navigation satellites to transmit signals to Earth — recorded between 2019 and 2026, the researchers built a triangulation formula that converts differences in signal intensity across affected antennas into a single geographic fix. The result points to a specific Russian spacecraft, according to Defense One.
That satellite is Cosmos 2546, launched in May 2020 as part of Russia's Edinaya Kosmicheskaya Sistema missile-warning constellation in a Molniya orbit. The attribution is the researchers' interpretation, drawn from the signal pattern and orbital geometry, not a confirmed act of Moscow-directed GPS attack. The paper, published this month in the journal Navigation, includes the team's own caveat: the bursts appear intentional but are too short to cause any real effect, and the researchers offer no specific theory about Russia's intentions.
What the team does offer is a reproducible method. Differential signal intensity across a continent-scale array rules out ground jammers and airborne platforms by sheer area. The same approach could, in principle, be applied to future interference events, giving insurers, aviation regulators, and diplomats something more durable than a fresh anecdote about GPS going dark.
The geographic footprint — Romania to Greenland — signals continent-scale intermittent interference rather than localized jamming. Defense One frames the result as a possible qualitative escalation in GNSS interference, language the underlying peer-reviewed work has not been independently confirmed to use.
For PNT resilience work, the practical question is whether the dataset and method survive peer review and replication. The team says the math is straightforward; the harder part is sustaining the monitoring antenna network required to keep generating comparable fixes. Independent GNSS-incident reporting from outlets such as GPS World, the European GNSS Agency, or C4ADS has not yet been cross-checked against the 75-event figure.
What changes now is what counts as evidence. Until this paper, a 10-second GPS blip was noise to be filtered. After it, the same blip is a data point. The story to watch next is the paper's full release: the exact beam footprint model, the countermeasures the authors propose, and whether independent teams can reproduce the 75-event record. Until then, Cosmos 2546 sits in a category the field has not had a name for: a spacecraft that may be disrupting GPS without intending to.