--- title: "When the GPS Signal Dies: Rubidium Clouds, Diamond Defects and 27 Metres Off Greenland" description: "With jamming now traced to Russian satellites, quantum sensors and stray radio waves are becoming credible backups" author: "Inez Holloway" published: 2026-09-27T10:25:46Z modified: 2026-09-28T09:18:25Z url: https://rews.cc/a/when-the-gps-signal-dies-rubidium-clouds-diamond-defects-and-1b1aca language: en tags: ["ai", "quantum", "gps", "drones", "space", "tech"] publisher: "Rews (https://rews.cc)" --- # When the GPS Signal Dies: Rubidium Clouds, Diamond Defects and 27 Metres Off Greenland *With jamming now traced to Russian satellites, quantum sensors and stray radio waves are becoming credible backups* By Inez Holloway · September 27, 2026 · https://rews.cc/a/when-the-gps-signal-dies-rubidium-clouds-diamond-defects-and-1b1aca ## In brief - June paper led by UT Austin’s Todd Humphreys ties European GPS disruptions since 2019 to Russian satellites - Q-CTRL’s quantum magnetometer beat a leading GPS alternative on a trial flight over New South Wales last year - Sparse, inconsistent crustal maps are MagNav’s main obstacle; SandboxAQ wants a joined-up global mapping effort - ASPIN Lab’s Matrix placed a Greenland cruise ship within 27 metres using 21 Starlink and OneWeb satellites - Psiaki’s method — broadcast departure times plus signals from four Starlink satellites — could fix a position within two metres It costs little to drown out a satellite navigation signal, and so, when the shooting starts, that signal is among the first things to stop arriving. Stronger signals from newer satellites make the jamming harder. What has changed is the scale: in June, a paper posted by a team led by Todd Humphreys of the University of Texas at Austin traced scores of broad GPS disruptions, across Europe and elsewhere since 2019, to a constellation of Russian satellites. Jamming from orbit — or even inadvertent interference from it — could, the research suggests, touch an entire continent. The Economist, which reported the finding this week alongside a survey of the alternatives, divides the escape routes in three. Mount cameras on a craft and track the terrain below: it works only over land, and fails against cloud, haze or darkness. The other two are stranger. One reads the earth’s magnetic crust through quantum phenomena — “the angular momentum of photons and the spin of electrons.” The other fixes position from ambient radio waves broadcast for other purposes. Neither matches the precision of satellite navigation yet. Both are becoming credible backups. Many rocks are weakly magnetic, and their irregular distribution leaves the planet’s field blotched with anomalies, detectable under almost any conditions and as easily at sea as over land. A plane carrying a sensitive enough magnetometer, and software that matches the readings to a map, can plot where it is. Last year an Australian firm, Q-CTRL, flew such a magnetic navigation system over New South Wales and, in its trial flight, outperformed a leading GPS alternative. Q-CTRL’s instrument is an optically pumped magnetometer. A laser is fired into a cloud of rubidium atoms; tuned correctly, the atoms’ electrons enter a state extremely sensitive to external magnetic fields. A second laser tracks that behavior, and from the numbers the software derives both the field and the place. Others are betting that the direction of the field matters as much as its strength. SBQuantum, a Canadian company, builds a magnetometer with a diamond at its heart. Diamonds are not normally magnetic, until the lattice is broken in a particular way: two of its carbon atoms are replaced by a single atom of nitrogen, leaving a gap — a nitrogen-vacancy centre — around which electrons gather. The regularity of the crystal then becomes a three-dimensional grid against which the electrons’ behavior can be read, and the field measured in vector form for sharper maps. The catch is the maps themselves. Good maps of the crustal field are scarce; the ones that exist are inconsistent with each other and often not accurate enough to navigate by. SBQuantum therefore flies its own survey missions. SandboxAQ, a Californian company, argues instead for improving and joining the maps that already exist, and in June submitted a paper proposing that America’s government agencies, armed forces and researchers do so together, on a global scale. They may soon get their wish: that same month, Donald Trump called for advances in quantum sensing and ordered the Department of War to explore applications. The second school listens rather than measures. America’s Federal Communications Commission has pressed for more work on these “signals of opportunity,” and one technique is the Doppler shift — the rising, then falling, frequency of a satellite passing overhead, the whistle of a train. The closer the observer to the satellite’s path, the faster the change. Combine published orbits with the rate of that shift and a position emerges. The demonstration was almost cinematic. On August 24, 2024, aboard a cruise ship off western Greenland, a team from ASPIN Lab at Ohio State University had their system, called Matrix, process transmissions from twenty-one Starlink and OneWeb broadband satellites passing overhead. Given only a starting position and speed, Matrix placed the ship twenty-seven metres from its true position. Doppler has a known weakness: satellites drift off course, and they are deliberately manoeuvred to avoid collisions, which throws off positions estimated from their signals. Stefano Tomasin of the University of Padua says errors of between 100 metres and more than a kilometre are common. The fixes are arriving from two directions. SpaceX now publishes predictions for the orbits of its more than 11,000 Starlink satellites — predictions that account for planned manoeuvres, which the tracking data produced by America’s Space Force do not — and Humphreys says the forecasts tend to be accurate to within a couple of metres of where the satellites turn out to be. And Mark Psiaki of Virginia Tech, funded by America’s Department of Transportation, is building a master receiver at a known location to compare expected Doppler shifts with observed ones and broadcast corrections to receivers at unknown locations; he hopes it will yield positioning accurate to about ten metres. The same master receiver enables a second trick: timing. Knowing where the master sits, and where SpaceX predicts a satellite to be, it can be worked out when a signal must have been transmitted, and the departure time broadcast onward. Each receiver then measures how long the same signal took to arrive, and therefore its distance from the satellite. With distances from four Starlink satellites, plus Doppler data, Psiaki says, a receiver should be able to fix itself within just two metres. The quest has extended to cell towers. In tests with America’s air force, ASPIN Lab has navigated an aircraft by tracking tiny shifts in mobile-phone signals, and Matrix has been licensed to companies and to military users. Zak Kassas, the lab’s director, expects the kit ready for market within two years. The maps will be drawn. Whatever the market decides, the number from that August stands: twenty-seven metres.