What Is GNSS?

How a receiver actually finds its coordinates

The whole thing rests on a simple physical principle. A satellite sends a signal, the receiver measures how long it took to arrive, and multiplies that time by the speed of light. That gives the distance to that particular satellite.

One satellite is nowhere near enough, and neither are two or three. A full three-dimensional position, meaning latitude, longitude and altitude, only comes through once the receiver locks onto at least four satellites at the same time. From there it is simple math: the more satellites in view of the antenna, the more stable the result. This matters most in places where buildings or dense tree cover block part of the sky.

The whole system rests on three parts. The space segment is the satellites themselves, orbiting Earth and constantly broadcasting signals. The ground segment is a network of monitoring stations that track those orbits and correct the satellites' positions. And the user segment covers everything that receives the signal, from an ordinary smartphone to a field-grade geodetic station.

Four systems, each with its own history

The Americans launched GPS back in 1978, and it became the de facto standard people associate with satellite navigation in general. Russia developed GLONASS in parallel, following a similar constellation logic. Europe's Galileo came later and was built with civilian accuracy as a priority. China's BeiDou finished rolling out full global coverage in 2020, and today it counts as one of the largest constellations by satellite number.

System	Operator	Number of satellites	Approximate accuracy
GPS	United States	about 31	5-10 m
GLONASS	Russia	24	5-10 m
Galileo	European Union	24-30 (target constellation)	up to 1 m in public mode
BeiDou	China	35	under 10 m

Hardly anyone relies on a single constellation alone anymore. Multi-GNSS receivers pick up signals from several systems at once, and that is exactly what saves the day in dense urban areas or forests, where part of the satellites simply aren't visible from a given spot.

Why accuracy jumps from meters to centimeters

Without correction

A standard receiver, with no additional corrections, gives an error of roughly 5 to 10 meters. That is more than enough for a car navigator or a hiking track. But once the job calls for precision down to a single centimeter, that level of accuracy simply won't cut it, and something else is needed.

With correction

SBAS pulls accuracy up to 1-3 meters by broadcasting real-time corrections through additional augmentation satellites. The real jump, though, comes from RTK. The receiver compares its own signal against data from a base station, and the error drops to just a few centimeters. It's worth reading up separately on how RTK actually works, since this is the technology behind modern precision farming and most surveying jobs.

Where it actually gets used

The list of applications moved well past car navigation a long time ago.

road, marine and aviation navigation;
surveying and construction layout work;
precision farming: guiding a tractor along parallel lines, applying fertilizer and spraying without overlaps or skipped strips;
monitoring deformation in bridges and dams;
logistics and real-time fleet tracking.

Every one of these cases comes down to the same choice. Basic GNSS accuracy works fine for general orientation, but wherever the cost of a mistake is measured in centimeters of a field or a construction line, there's no getting around RTK or some other form of signal correction.