Multilateration (MLAT): Tracking Planes That Don't Want to Be Found

Somewhere over the English Midlands, a business jet is trying to keep a low profile. Its transponder is switched on, because flying without one in controlled airspace is not an option, but it broadcasts no ADS-B position. On a basic receiver it is a voice without a location, a callsign muttering in the dark. And yet on the tracking map its icon slides along smoothly at 41,000 feet, position updating every few seconds. No radar painted it. No satellite fixed it. Four ordinary rooftop antennas simply heard the same radio message at four minutely different instants, and geometry did the rest. This is multilateration, MLAT to its friends, and it is the closest thing flight tracking has to a magic trick.

The core idea: Time is distance

Radio waves travel at the speed of light, roughly 300 meters every microsecond, or about 30 centimeters every nanosecond. That constancy turns arrival times into a ruler. Imagine an aircraft transmits a single Mode S reply. A receiver 60 kilometers away hears it 200 microseconds after transmission; a receiver 90 kilometers away hears it 300 microseconds after. No individual receiver knows when the message left the aircraft, so absolute times are useless on their own. But the difference between arrival times at two receivers, 100 microseconds in this example, tells you the aircraft is exactly 30 kilometers closer to the first receiver than to the second.

The set of all points that are 30 kilometers closer to one location than another forms a curved surface, a hyperboloid. Every pair of receivers that hears the message contributes one such surface. Two pairs narrow the aircraft down to the curve where their surfaces intersect; three pairs pin it to a point. That is why the technique's formal name is time difference of arrival, TDOA, and why the practical minimum is four receivers for a full three-dimensional fix. In practice, trackers often use the aircraft's own Mode C altitude as the vertical coordinate, which lets three well-placed receivers produce a usable two-dimensional position.

If this sounds familiar, it should. GPS is the same geometry running in reverse: satellites broadcast timed signals and your phone computes where it must be to explain the arrival times. MLAT flips the arrangement, with the mystery transmitter in the sky and the synchronized listeners on the ground.

The tyranny of the nanosecond

The catch is synchronization. A timing error of one microsecond between two receivers corrupts the geometry by 300 meters, so the receivers' clocks must agree to within tens of nanoseconds, a precision no consumer quartz oscillator can hold. Community tracking networks solve this in two main ways. Some stations discipline their clocks with GPS timing pulses, which are good to a few tens of nanoseconds. Others use reference transmissions: when a receiver hears an ADS-B aircraft whose exact position is known from its own broadcast, the expected arrival time can be computed and the receiver's clock error measured against it, a continuous self-calibration using the sky itself as the reference.

Then comes the computation. The central server must recognize that the same 64-microsecond Mode S reply was heard by stations in four different towns, match those detections among millions of messages per second, solve the hyperbolic equations, and filter the jitter, all in near real time. It is a testament to modern commodity computing that volunteer networks do this continuously for thousands of aircraft. The result is typically accurate to a few hundred meters in well-covered areas, coarser than ADS-B's GPS-grade fixes but easily good enough to draw an honest track.

Who shows up on MLAT, and why

MLAT works on any aircraft that transmits anything on 1090 MHz, because the technique needs only the signal's timing, not its content. In practice, the aircraft that appear on tracking maps courtesy of MLAT fall into a few recognizable families. There are older airliners and cargo aircraft with Mode S transponders that predate ADS-B mandates. There are business jets whose owners chose not to broadcast position, sometimes for privacy, sometimes because their equipment qualified for exemptions; our companion piece on what is visible when you track a private jet covers the cat-and-mouse in detail. There are general aviation aircraft below the mandate floor. And there are military transports and tankers that fly with Mode S on for safety in civil airspace but position reporting off.

Coverage, not aircraft type, is the real gatekeeper. A message must be heard by at least three or four stations simultaneously, so MLAT thrives where receivers are dense, across Europe, Japan, the coastal United States, and fades where they thin out. Over oceans it is impossible for community networks, since there are no rooftops; that gap belongs to space-based ADS-B receivers riding on satellites. Every receiver added at the edge of a network expands the MLAT envelope, which is one of the best reasons to set up your own Raspberry Pi receiver if you live outside a big city; rural stations are disproportionately valuable.

MLAT needs nothing from the aircraft but the sound of its voice. As long as a transponder speaks, four listeners with good clocks can find the speaker.

The professional cousin: WAM

Multilateration is not just a hobbyist workaround; it is certified surveillance infrastructure. Air navigation service providers deploy wide area multilateration, WAM, as a radar replacement or supplement, with engineered ground stations, redundant timing and guaranteed accuracy. Austria uses WAM across mountainous terrain where radar coverage is patchy and new radar heads would be ruinously expensive. The Czech Republic ran one of Europe's pioneering national WAM systems. In the United States, the FAA has used WAM in Colorado's high terrain, where mountain shadowing blinds conventional radar over ski-country airports. Eurocontrol publishes specifications for WAM performance, and ICAO recognizes multilateration as a standard surveillance technique alongside radar and ADS-B.

WAM installations also solve airport surface surveillance. A dozen small receivers around a terminal can track every transponder-equipped aircraft and vehicle on taxiways to within a few meters, feeding the runway-incursion warning systems at major hubs. The same mathematics that finds a coy business jet at 41,000 feet also notices a catering truck wandering toward an active runway. You can browse the world's largest airports, many of which run exactly this kind of system, in our airport directory.

Reading MLAT tracks on the map

Once you know what to look for, MLAT targets are easy to spot on a live tracking map. Their positions update a little less often and a little less smoothly than ADS-B traffic. Their listed position source will say MLAT rather than ADS-B on services that disclose it. Altitude usually comes from the aircraft's own Mode C or Mode S reply, so it remains crisp even when the horizontal fix wobbles. Tracks may cut out at low altitude as the aircraft descends below the shared coverage of enough stations, then reappear on final approach near a well-instrumented airport.

There is a quiet philosophical point in all this. Aviation's surveillance layers form a ladder of cooperation: primary radar needs nothing from the aircraft, MLAT needs only its voice, secondary radar needs an answer, and ADS-B needs the whole confession, as we chart in our comparison of primary and secondary radar and our deep dive into how ADS-B works. MLAT sits at the sweet spot for trackers: it demands no honesty about position, only the unavoidable act of transmission. Silence remains the only true invisibility, and in controlled airspace, silence is not on the menu.

The business jet over the Midlands never did say where it was. It did not need to. Four strangers' antennas, a shared clock and some seventeenth-century geometry said it for them, which is a fine thing to remember the next time someone tells you that you cannot find what refuses to be found.