RTK positioning principles
Real-Time Kinematic positioning uses GNSS carrier-phase measurements. A station at known coordinates observes common satellites and provides information that reduces shared errors. The rover combines corrections with its own observations and attempts to resolve integer ambiguities. A validated integer solution is commonly reported as Fix.
Float uses unresolved ambiguities and has variable accuracy. Single generally describes autonomous positioning. These labels are operational indicators; geometry, multipath, antenna setup, correction age, reference frame, and independent field checks still matter.
Baseline distance influences error correlation. A network of stations can model spatial atmospheric effects and provide network corrections over a wider region.
RTK network architecture
A local service can begin with one base and a caster. Regional services collect several stations into a processing engine. The caster publishes raw station streams and network products, while the control center shows which bases and rovers are active.
Resilience covers stations, links, servers, DNS, certificates, accounts, and sourcetables. A second executable without synchronized configuration is not a complete failover system.
NRTK, VRS, MAC, and FKP
Network RTK estimates spatial errors from multiple stations. VRS creates a stream that behaves like a virtual station near the rover’s GGA position. MAC provides master and auxiliary information, while FKP represents area correction parameters.
VRS is session-specific and usually requires GGA. A single shared upstream relay may therefore be unsuitable for multiple distant rovers. Imported mountpoints must be classified by their actual behavior.
A nearest-station service reduces baseline distance but does not perform the same network modeling. Clear mountpoint descriptions prevent users from confusing the products.
GPS, GLONASS, Galileo, BeiDou, and SBAS
Multi-constellation tracking improves satellite availability and geometry, especially near obstructions. The base must actually transmit the corresponding RTCM messages, and the rover must decode them.
MSM variants provide different levels of detail and bandwidth. Message choice and interval are validated against real receivers and network conditions. More data is not automatically better when clients cannot use it.
Consistent antenna models, reference coordinates, and processing conventions remain as important as the number of satellites.
Coverage, baseline, and accuracy
Commercial coverage is not just a circle. Atmosphere, terrain, mobile connectivity, station density, and sky visibility change performance. Service areas should be supported by field tests of initialization time, repeatability, Fix rate, and known-point residuals.
The control center calculates base-rover distance and can enforce a station radius. A margin is prudent because GGA positions may be rounded or delayed.
Urban multipath can dominate the error even near a base. Good antennas, clear sky, correct heights, and professional procedures remain necessary.
Quality of service and use cases
Quality combines caster uptime, station availability, correction age, authentication success, and rover outcomes. Tenant- and mountpoint-level indicators reveal local issues hidden by global averages.
Surveying, construction, agriculture, machine control, drones, and mapping have different continuity and accuracy needs. RINEX archives provide a PPK fallback for missions where live corrections were incomplete.
| Symptom | Possible causes | Checks |
|---|---|---|
| No connection | DNS, port, TLS, credentials | Resolution, firewall, response, logs |
| Data but no Fix | Incomplete RTCM, baseline, sky, frame | Messages, GGA, distance, control point |
| Repeated disconnects | Mobile network, timeout, client | End reason, interval, User-Agent |
| Coordinate offset | Base coordinates, frame, antenna | 1005/1006, station log, known point |
RTK and NRTK FAQ
Is there a universal maximum RTK baseline?
No. Acceptable distance depends on atmospheric conditions, station density, signals, terrain, communications, and the required accuracy. Local validation is essential.
What is the difference between RTK and NRTK?
RTK uses real-time corrections, often from one base. NRTK uses a network of reference stations to model spatial errors over a wider service area.
Why can a rover remain Float?
Possible causes include obstructions, multipath, stale or incomplete corrections, a long baseline, poor satellite geometry, incompatible messages, or insufficient initialization.