Introduction
"High-precision GNSS relies on a cooperative system: a base station and one or more rovers."
Single-receiver GNSS positioning, no matter how sophisticated, is fundamentally limited by errors that cannot be resolved with one antenna alone - satellite clock errors, orbit errors, ionospheric delay, and tropospheric delay all remain at levels of metres without external reference. The base-and-rover architecture solves this by introducing a second receiver at a known location, transforming the positioning problem from an absolute one into a differential one, and achieving centimetre-level accuracy.
The Base Station: The Foundation of Accuracy
The base station is a GNSS receiver installed at a point whose coordinates are precisely known. It collects observations continuously and generates correction data - typically in RTCM (Radio Technical Commission for Maritime Services) format - that describe the difference between the satellite-derived measurements and the known geometry at that point. These corrections encapsulate the combined effect of atmospheric delays, satellite clock errors, and orbit errors as seen at the base location.
The accuracy of the base station's known position is the ceiling on the accuracy the rover can achieve. If the base is placed at a position that is wrong by 10 cm, every rover connected to it will also be wrong by approximately 10 cm. Establishing a base on a trusted geodetic control point, or deriving its coordinates through a careful static GNSS survey tied to a national network, is therefore the first and most critical step in any precision GNSS deployment.
Key base station requirements include:
- A stable, permanent, or well-benchmarked mounting point with a clear sky view above the elevation mask
- A high-quality, survey-grade antenna - ideally a choke ring - to minimise multipath and phase center variation
- Continuous, uninterrupted power supply and data logging
- A reliable communication link to transmit corrections to rovers
The Rover: The Mobile Measurement Tool
The rover is the mobile receiver that moves through the area of interest making measurements. It receives real-time corrections from the base station and uses them to remove the shared atmospheric and clock errors from its own measurements. By differencing its observations against the base's, the rover can resolve the integer ambiguities in its carrier phase measurements and achieve RTK Fixed status - the centimetre-accurate positioning mode.
The rover's position is always computed relative to the base station. This is why RTK accuracy specifications are quoted as a baseline-dependent value - for example, 8 mm + 1 ppm, where the 1 ppm term means accuracy degrades by 1 mm per kilometre of separation from the base. This degradation occurs because the atmospheric conditions experienced by the base and rover diverge as the distance between them increases.
Correction Flow: From Base to Rover
The correction data flow follows a well-defined path. The base receiver computes RTCM correction messages every second (or more frequently) and transmits them via a communication link. The rover receives these corrections, applies them to its own measurements in real time, and solves for its position. The complete cycle - observation, correction generation, transmission, reception, and position computation - must complete within one to two seconds for effective RTK operation.
RTCM correction messages carry satellite observations, satellite positions, and atmospheric model parameters. The rover's engine uses these to form double-difference carrier phase equations - differencing between satellites and between receivers - that cancel receiver and satellite clock errors and dramatically reduce atmospheric errors.
Fixed Baseline vs Moving Baseline
Fixed Baseline RTK
The standard configuration: the base is stationary at a known point and the rover moves freely within the operational area. This is the architecture used in surveying, construction layout, precision agriculture, and most GIS data collection. The baseline - the vector between base and rover - changes continuously as the rover moves, but the base position remains fixed.
Moving Baseline RTK
In moving baseline mode, both the base and rover receivers are in motion, but they are physically attached to the same platform - typically a vessel, aircraft, or vehicle. The relative vector between the two antennas is fixed by their rigid mounting. Processing this fixed-length, orientation-varying baseline gives the heading and attitude of the platform with centimetre-level relative accuracy. This technique is widely used in marine navigation, autonomous vehicles, UAVs, and antenna stabilisation systems.
| Architecture | Base Movement | Primary Output | Typical Application |
|---|---|---|---|
| Fixed Baseline RTK | Stationary | Rover absolute position | Surveying, construction, agriculture |
| Moving Baseline RTK | Moving (same platform) | Heading and attitude | Marine, UAV, autonomous vehicles |
| Network RTK (CORS) | Virtual (networked) | Rover absolute position | Wide-area precise positioning |
System Design Considerations
Precision GNSS system performance depends more on the overall system design than on any individual component. A high-end rover paired with a poorly placed base will perform worse than a mid-grade rover connected to a well-established network reference station. Engineers must design the system as a whole, considering base placement, antenna selection, correction communication, power reliability, and rover configuration together.
Vital Points
- The base station defines the accuracy ceiling - a wrongly coordinated base produces systematically incorrect rover positions across the entire project.
- The rover depends entirely on the corrections it receives; without a reliable correction link, RTK degrades to standalone GNSS accuracy.
- Baseline length matters - accuracy degrades with distance from the base due to atmospheric divergence.
- System design as a whole matters more than individual component specifications.