How GPS Works: Satellites, Trilateration, and Atomic Clocks

Navigation by Light Speed

Your phone knows where you are to within a few meters because it is measuring the travel time of radio signals moving at the speed of light. Light covers roughly 30 centimeters in one nanosecond. To achieve meter-level positioning accuracy, a GPS receiver must measure signal travel times to nanosecond precision — timing accuracy that requires atomic clocks on satellites, general relativistic corrections, and a receiver algorithm simultaneously solving for four unknowns: latitude, longitude, altitude, and receiver clock error.

The Global Positioning System (GPS) was developed by the United States Department of Defense, declared fully operational in 1995, and made freely available to civilian users in that year. Selective Availability — intentional degradation of civilian signal accuracy — was deactivated in May 2000, improving civilian accuracy from ~100 meters to ~15–30 meters with no hardware changes. Today, with correction services, commercial GPS achieves centimeter-level accuracy. Modern aircraft, precision agriculture equipment, autonomous vehicles, and time-synchronization systems for telecommunications networks all depend on GPS or its international equivalents.

The Satellite Constellation

GPS uses a constellation of at least 24 operational satellites (typically 31 are active) distributed across six orbital planes inclined at 55° to the equator, at an altitude of approximately 20,200 kilometers. This configuration ensures that at least four satellites are always visible from any point on Earth's surface at any time, with most locations seeing six to twelve satellites simultaneously.

Each GPS satellite broadcasts on two primary frequencies: L1 (1,575.42 MHz) and L2 (1,227.60 MHz). Civilian receivers primarily use L1; modern multi-band receivers use L1 and L5 (1,176.45 MHz) for enhanced accuracy. The signals carry navigation messages containing satellite ephemeris data (precise orbital position), satellite health status, and time information referenced to GPS time — maintained by the GPS Master Control Station at Schriever Space Force Base, Colorado.

Trilateration: Position from Distances

GPS positioning uses trilateration, not triangulation. Triangulation uses angles to determine position; trilateration uses distances. Each satellite transmits a coded ranging signal — a pseudorandom noise (PRN) code — that the receiver replicates internally. By measuring the delay between the received code and its local replica, the receiver computes the signal travel time and multiplies by the speed of light to get the distance to that satellite.

If perfect timing were available, three satellites would uniquely determine a 3D position (the intersection of three spheres). In practice, the receiver's clock is far less accurate than the satellites' atomic clocks, introducing a common timing error into all distance measurements. A fourth satellite measurement allows the receiver to solve for this clock error simultaneously with the three position coordinates, yielding four equations in four unknowns:

• d₁ = c × (t_received₁ − t_sent₁) + c × δt — distance to satellite 1 (true distance plus clock error term)
• d₂ = c × (t_received₂ − t_sent₂) + c × δt — distance to satellite 2
• d₃ = c × (t_received₃ − t_sent₃) + c × δt — distance to satellite 3
• d₄ = c × (t_received₄ − t_sent₄) + c × δt — distance to satellite 4

The receiver solves this system iteratively. Because the clock error is common to all measurements, the four measurements are self-consistent and allow precise determination of all four unknowns. Each additional satellite beyond four provides a redundant measurement that enables quality control and improves accuracy through least-squares estimation.

Atomic Clocks: The Core Technology

Each GPS satellite carries four atomic clocks (two cesium, two rubidium), providing redundancy and cross-checking. Rubidium clocks are smaller and lighter; cesium clocks are more stable over long periods. The operational clock switches automatically to the most accurate unit.

Cesium atomic clocks measure time using the transition frequency of cesium-133 atoms. When cesium atoms transition between two specific hyperfine energy levels, they emit or absorb microwave radiation at precisely 9,192,631,770 Hz — the SI definition of the second since 1967. An atomic clock is accurate to approximately 1 part in 10¹³, meaning it loses or gains less than one nanosecond per day. Over the 20-year life of a GPS satellite, such a clock would drift by less than one millisecond total. Ground control stations upload daily corrections to satellite clocks to keep them synchronized to within 20–30 nanoseconds of GPS time.

Relativistic Corrections: Where Einstein Meets Navigation

Without relativistic corrections, GPS would fail to provide useful positioning within hours of startup. Two separate relativistic effects apply:

• Special relativity (time dilation): Satellites move at ~3.9 km/s relative to Earth's surface. Moving clocks run slow by a factor of √(1 − v²/c²). At satellite velocity, GPS clocks run slow by approximately 7.2 microseconds per day compared to clocks on Earth's surface.
• General relativity (gravitational time dilation): Clocks at higher gravitational potential (farther from Earth's center) run faster. GPS satellites at 20,200 km altitude are in a weaker gravitational field than surface clocks. General relativistic effect adds approximately 45.9 microseconds per day — larger than the special relativistic effect and in the opposite direction.
• Net effect: GPS satellite clocks run fast by approximately 38.4 microseconds per day relative to surface clocks. Without correction, this timing error would translate to a positioning error of approximately 11 kilometers per day.

GPS system designers correct for this by setting satellite clock frequencies slightly slow before launch — at 10.22999999543 MHz instead of the nominal 10.23 MHz — so that in orbit, relativistic effects bring them back into synchrony with ground clocks.

Error Sources and Accuracy Enhancement

Differential GPS (DGPS) and Real-Time Kinematic (RTK) techniques use fixed reference stations at known locations to measure current GPS errors and broadcast corrections to nearby receivers. RTK achieves centimeter-level horizontal accuracy. Precise Point Positioning (PPP) uses globally distributed reference station networks to estimate satellite clock and orbital errors at the millimeter level, enabling centimeter positioning anywhere in the world without local reference stations — the standard for geodesy, precision agriculture autopilot systems, and autonomous vehicle HD mapping.