TLDR: RTK Corrections
RTK corrections are real-time data from ground-based reference stations that eliminate GNSS satellite signal errors, improving positioning accuracy from 3–5 meters to 1–2 centimeters. Base stations at precisely known locations measure errors, encode them as RTCM messages, and deliver them to rovers via NTRIP (internet), radio, or satellite. Single-baseline RTK works within ~30 km of a station, while Network RTK (VRS) uses multiple stations to extend coverage and maintain accuracy across larger areas. Correction quality depends on latency, distance to base stations, station hardware, and multi-constellation support. For production applications, commercial RTK services with uptime SLAs are recommended over free networks like state DOTs or RTK2GO.
RTK corrections are real-time data transmitted from ground-based reference stations to GNSS receivers that eliminate satellite signal errors, transforming meter-level positioning into centimeter-level accuracy. Where a standard GNSS receiver delivers location accuracy within 3–5 meters, applying RTK corrections refines this to 1–2 centimeters horizontally — a 100x improvement that makes the difference between “roughly here” and “exactly here.”
For professionals in surveying, construction, agriculture, autonomous systems, robotics, and the emerging world of Physical AI (autonomous systems that interact with the physical world) – understanding how RTK correction services work — and what affects their quality — is essential to selecting the right solution and getting the most out of precision positioning technology.
This guide explains what RTK corrections contain, how they’re generated and delivered, what affects their quality, and how to evaluate an RTK correction service for your application.
What Are RTK Corrections?
RTK corrections are supplemental data that account for errors in GNSS satellite signals. They are not a position — they are the difference between what a reference station observes and what it should observe based on its precisely known location. A rover receiver applies this difference to its own measurements, canceling out shared error sources and dramatically improving accuracy.
RTK works across all major GNSS constellations — GPS, GLONASS, Galileo, and BeiDou — and the term “RTK corrections” applies regardless of which constellation the receiver is tracking. You may encounter the legacy phrase “RTK GPS corrections,” but modern systems use multi-constellation GNSS for better satellite availability and positioning performance.
What makes RTK corrections fundamentally different from simpler correction methods like DGNSS is the measurement type they correct. Standard GNSS uses code-phase measurements based on the satellite signal’s pseudorandom noise code, which has a wavelength of roughly 300 meters. RTK uses carrier-phase measurements — the actual radio wave that carries the signal — with a wavelength of approximately 19 centimeters. This much finer measurement resolution is what enables centimeter-level positioning.
For a broader overview of RTK positioning — including how it compares to standard GNSS, when to use it, and how to choose a solution — see our companion guide: What is RTK?
What is RTK GPS?
“RTK GPS” refers to applying RTK corrections specifically to GPS signals. The term became widespread when GPS was the dominant satellite constellation, but modern RTK systems simultaneously process signals from GPS, GLONASS, Galileo, and BeiDou. Multi-constellation RTK provides access to more satellites at any given time, which improves accuracy, reliability, and performance in challenging environments like urban areas or under tree canopy.
Throughout this guide, we use “RTK corrections” to reflect current practice. If your documentation or hardware references “RTK GPS corrections,” the underlying technology is the sa
RTK Float vs. RTK Fixed: Understanding Fix Status
When a rover first begins receiving RTK corrections, it enters a convergence process with two distinct states:
RTK Float: Corrections are being applied, but the receiver has not yet resolved integer ambiguity — the mathematical challenge of determining the exact number of carrier-phase wavelengths between each satellite and the receiver. In Float mode, accuracy improves to the sub-meter level but has not yet reached full RTK precision.
RTK Fixed: The receiver has successfully resolved all integer ambiguities and achieved centimeter-level accuracy. This is the target state for professional applications — whether you’re surveying boundaries, guiding autonomous machinery, or conducting precision agriculture operations.
Convergence from Float to Fixed typically takes seconds with modern receivers and a quality correction stream, though environmental conditions, satellite geometry (PDOP), and receiver quality all influence convergence time.
How Accurate Are RTK Corrections?
RTK accuracy is one of the technology’s defining advantages:
Positioning Method | Horizontal Accuracy | Vertical Accuracy |
Standard GNSS (uncorrected) | 3–5 meters | 5–10 meters |
DGNSS | 0.5–1 meter | 1–2 meters |
RTK (Fixed) | 1–3 centimeters | 1-3 centimeters |
These figures represent RTK Fixed accuracy under good conditions — open sky, quality receiver, and proximity to a base station. In practice, accuracy degrades approximately 1–1.5 cm per 10 km of distance from the nearest reference station. For single-baseline RTK, the effective operating range is typically 30 km from a base station. Beyond that, atmospheric differences between the base and rover become too large for the corrections to fully compensate.
Network RTK extends this range significantly by using multiple stations to model atmospheric conditions across a region and generating a Virtual Reference Station (VRS) near the rover’s location. This keeps the effective baseline short even when the rover is far from any physical station.
For a detailed analysis of how distance affects RTK accuracy — including the underlying math — see Understanding RTK Accuracy Over Distance.
How RTK Corrections Work
Understanding the end-to-end correction pipeline — from satellite signal to centimeter-accurate position — helps you select the right service, troubleshoot issues, and set appropriate expectations for your application.
GNSS Satellite Signals and Error Sources
GNSS satellites broadcast timing signals that receivers use to calculate distance (pseudorange) to each satellite. With signals from four or more satellites, the receiver can compute its position. However, several error sources degrade this calculation:
Ionospheric delay: Charged particles in the upper atmosphere slow the signal, introducing distance errors of 1–50 meters depending on solar activity and time of day.
Tropospheric delay: Water vapor and temperature variations in the lower atmosphere cause additional signal delays, typically adding 2–25 meters of error.
Satellite orbit and clock errors: Despite regular updates (ephemeris data), satellite positions and onboard clocks have residual errors that contribute to positioning inaccuracy.
Multipath: Signals reflecting off buildings, terrain, or other surfaces before reaching the antenna create false distance measurements.
These errors are spatially correlated — two receivers in the same geographic area experience similar atmospheric and orbital errors. This spatial correlation is the fundamental principle that makes RTK corrections work: a base station at a known location can measure these shared errors and provide the delta that a nearby rover needs to cancel them out.
Base Stations and Reference Networks
A base station (also called a reference station) is a GNSS receiver installed at a precisely surveyed location. Because its true position is known to within millimeters, the station can continuously compare its observed satellite measurements against what it should observe, isolating the errors affecting signals in that area.
The computed corrections are generated for each satellite signal the station tracks. These corrections represent the combined effect of atmospheric, orbital, and clock errors at that location and time.
Single-baseline RTK uses corrections from one physical base station. The rover connects to the nearest station and applies its corrections directly. This approach works well within approximately 30 km of the station, but accuracy degrades with distance as atmospheric conditions diverge between the base and rover locations.
Network RTK uses observations from multiple interconnected base stations across a region. The network models atmospheric and orbital errors spatially and generates a Virtual Reference Station — a synthetic set of corrections optimized for the rover’s specific location. This maintains short effective baselines across the entire coverage area, providing consistent accuracy without requiring proximity to any single physical station.
The quality of Network RTK corrections depends directly on the density of the underlying station network. Denser networks capture finer spatial variations in atmospheric conditions, which produces more accurate VRS interpolation, faster convergence, and more stable Fixed solutions. This is why two networks advertising “centimeter accuracy” can deliver meaningfully different field performance.
Correction Encoding: RTCM Format
RTK corrections are encoded using the RTCM (Radio Technical Commission for Maritime Services) standard — specifically RTCM 3.x, which defines the message format for carrier-phase observation data, station coordinates, and related metadata.
Key RTCM message types for RTK include observation messages for each constellation (GPS, GLONASS, Galileo, BeiDou) and station coordinate messages that define the reference station position. The receiver applies these correction observations alongside its own measurements to resolve integer ambiguities and compute a centimeter-accurate position.
The distinction between OSR and SSR correction encoding is covered in the comparison section below, as it directly affects how different correction methods work and which receivers they’re compatible with.
How Corrections Are Delivered
NTRIP (Internet-Based Delivery)
NTRIP (Networked Transport of RTCM via Internet Protocol) is the industry-standard protocol for delivering RTK corrections over the internet. The protocol uses a three-component architecture: an NTRIP server at the base station, an NTRIP caster that distributes correction streams, and an NTRIP client (your rover) that connects and receives data.
To connect, a rover needs a caster URL, port number, mountpoint name, and authentication credentials — typically provided by your RTK correction service. NTRIP requires a stable internet connection (cellular, WiFi, or ethernet) and delivers corrections with minimal latency, typically under one second.
For a complete walkthrough of NTRIP setup, architecture, and troubleshooting, see our What is NTRIP? guide.
Radio Link (Direct Delivery)
UHF or VHF radio modems can transmit corrections directly from a local base station to a rover without any internet infrastructure. This approach is common in traditional survey workflows where a team deploys their own temporary base station. Radio delivery is limited by line-of-sight requirements and range (typically under 10 km), but it works in remote areas without cellular coverage.
L-Band Satellite Delivery
L-Band delivery broadcasts corrections via geostationary communication satellites, eliminating the need for cellular or internet connectivity. This is particularly valuable for agriculture, maritime, and remote operations where reliable corrections are needed across large areas without cellular infrastructure.
The Correction Process: From Base to Rover
The complete RTK correction pipeline works as follows:
- Base stations observe: Reference stations continuously receive satellite signals and record carrier-phase observations across all tracked constellations.
- Errors are computed: Each station compares its observations to what it should measure based on its precisely known position, isolating atmospheric, orbital, and clock errors.
- Corrections are encoded: Error data is packaged as RTCM-format correction messages.
- Corrections are delivered: Messages are transmitted to the rover in real time via NTRIP, radio, or satellite.
- Rover resolves and positions: The rover combines corrections with its own observations, resolves integer ambiguities, and outputs a centimeter-accurate position.
What Is an RTK Correction Service?
An RTK correction service is a subscription-based solution that provides access to correction data from a professionally managed network of base stations. Instead of deploying, surveying, and maintaining your own reference station, you connect your RTK-enabled receiver to the service and receive corrections immediately.
Benefits of Using an RTK Correction Service
Managed RTK correction services handle the infrastructure complexity so you can focus on your application. Key advantages include no capital investment in base station hardware, no need to survey and maintain reference station locations, coverage across large geographic areas (regional, national, or global depending on the provider), professional uptime guarantees and monitoring, automatic network expansion as the provider adds stations, and compatibility with standard RTCM 3.x and NTRIP protocols.
For teams evaluating whether to build their own RTK infrastructure versus subscribing to a service, see Build Your Own RTK for a detailed comparison.
Free vs. Paid RTK Correction Services
Not all RTK correction services are equal, and the distinction between free and paid options matters for production applications.
Free services — including state DOT CORS networks, community casters like RTK2GO, and some university networks — provide access to correction data at no cost. These can work well for testing, prototyping, and hobbyist use. However, they typically offer limited coverage, no service level agreements (SLAs), variable reliability, and minimal technical support.
Paid commercial services provide guaranteed uptime (99.9%+), broader and denser station networks, customer support, developer tools, and the reliability that production-critical applications require. When your autonomous fleet, construction machine control system, or survey workflow depends on corrections being available, a commercial service with explicit reliability commitments is the appropriate choice.
For a comprehensive evaluation of NTRIP and RTK service providers, see our NTRIP / RTK Service Provider Buyer’s Guide.
What Affects RTK Correction Quality?
Not all RTK corrections deliver the same results. Several factors determine whether your receiver achieves a stable Fixed solution with optimal accuracy — and understanding these helps you evaluate services and troubleshoot field issues.
Correction Age (Latency)
RTK corrections are time-sensitive. The freshness of correction data — often called “correction age” — directly impacts positioning quality. Most RTK receivers can use OSR corrections up to approximately 60 seconds old, but performance degrades as latency increases. For optimal results, corrections should arrive within 1–2 seconds of being computed. Internet latency, cellular network congestion, and server processing all contribute to correction age.
Baseline Distance
Corrections computed at one location become less representative of conditions at another location as distance increases. For single-baseline RTK, accuracy degrades by roughly 1–1.5 cm per 10 km. Beyond 30 km, atmospheric differences between the base and rover may prevent the receiver from achieving or maintaining a Fixed solution. Network RTK mitigates baseline distance limitations by creating virtual corrections near the rover’s actual position, regardless of how far it is from any physical station.
Station Quality and Calibration
The quality of the base station hardware, its installation, and its ongoing maintenance directly affect correction accuracy. Geodetic-grade antennas with calibrated phase-center models, stable monuments that resist settling and thermal expansion, and regular calibration verification all contribute to reliable corrections. Stations built, deployed, and monitored by the correction provider tend to deliver more consistent quality than networks assembled from third-party or crowd-sourced stations.
Multi-Constellation Support
Corrections that include observations from all four major constellations — GPS, GLONASS, Galileo, and BeiDou — enable the rover to track more satellites simultaneously. More satellites means better geometry, faster convergence to Fixed, and more robust performance in environments with partially obstructed sky views. Base stations that only track GPS and GLONASS leave significant satellite resources on the table.
RTK Correction Limitations and Challenges
While RTK corrections deliver exceptional accuracy, understanding their limitations helps set appropriate expectations and plan for edge cases.
Coverage and Base Station Density
RTK corrections are only as good as the network generating them. In areas with sparse station coverage, effective baselines become long and accuracy suffers. Some providers have excellent coverage in one region but significant gaps in others. Before committing to a service, verify coverage density in your specific operating area — not just whether a provider claims coverage there.
Signal Obstructions and Challenging Environments
RTK requires clear sky visibility to enough GNSS satellites. Urban canyons (tall buildings on both sides of a street), dense tree canopy, and indoor or underground environments can reduce the number of visible satellites below what’s needed for a Fixed solution. In these challenging scenarios, many teams pair RTK with an inertial navigation system (INS) that uses an IMU to maintain positioning through brief GNSS outages via dead reckoning. Point One’s Positioning Engine provides this sensor fusion capability. For more on the underlying concepts, see Absolute vs. Relative Navigation.
Internet Connectivity Requirements
NTRIP-based correction delivery requires a stable internet connection. In remote areas where cellular coverage is unreliable, corrections may be intermittent. Solutions include L-Band satellite delivery for areas without cellular coverage, local base stations with radio link for targeted deployments, and store-and-forward approaches for post-processed workflows (PPK).
RTK vs. Other GNSS Correction Methods
RTK is one of several techniques for improving GNSS accuracy. Understanding the differences — including how corrections are represented and delivered — helps you choose the right approach for your application.
OSR vs. SSR: Two Ways to Represent Corrections
OSR (Observation Space Representation) delivers corrections as composite observations — all error sources lumped together — for the rover’s specific location. This is the traditional approach used by single-baseline RTK, Network RTK, and VRS services. OSR is straightforward, widely compatible with any RTK receiver, and places the computational burden on the correction provider rather than the rover. The trade-off: because corrections are location-specific, each user needs a unique data stream, which limits scalability.
SSR (State Space Representation) models each error source separately — satellite orbits, clocks, ionospheric delay, tropospheric delay — and sends those model parameters to the rover, which reconstructs its own correction. Because the same broadcast stream works for all users regardless of location, SSR scales efficiently to millions of devices. The trade-off: the rover needs more sophisticated processing firmware to interpret the individual error components.
A note on VRS and modeling: Network RTK services that use Virtual Reference Stations (VRS) do involve atmospheric modeling — the server interpolates ionospheric and tropospheric conditions across the network to synthesize observations for the rover’s location. But the model runs server-side; the rover still receives composite observations (OSR). This is different from SSR, where the model parameters themselves are sent to the rover.
For most users, this distinction matters when evaluating two things: receiver compatibility (OSR works with virtually any RTK receiver; SSR requires specialized firmware) and scalability (OSR requires a unique stream per user; SSR can broadcast one stream to all).
Comparison of GNSS Correction Methods
DGNSS | Single-Baseline RTK | Network RTK (VRS) | Virtual Corrections (SSR-modeled, OSR-delivered) | PPP | PPP-RTK | |
Representation | OSR | OSR | OSR | SSR on backend → OSR to rover | SSR | SSR |
Correction type | Code-phase differential | Carrier-phase from nearest physical base | Carrier-phase from VRS | Standard RTCM (looks like single-baseline) | Precise orbit + clock products | Orbit + clock + regional atmospheric models |
Uses atmospheric modeling? | No | No (relies on spatial correlation) | Yes (server-side interpolation) | Yes (full SSR model) | Orbits + clocks only | Yes (regional ionosphere + troposphere) |
Accuracy | 0.5–1 m | 1–3 cm | 1–3 cm | ~10 cm | 5–20 cm (post-convergence) | 2–10 cm |
Convergence | Instant | Immediate | Seconds | ~30 seconds | 15–30+ minutes | Seconds |
Receiver requirements | Any GNSS receiver | Dual-band RTK receiver | Dual-band RTK receiver | Dual-band RTK receiver | Specialized PPP firmware | Specialized PPP-RTK firmware |
Coverage | Near base station(s) | ~30 km from base | Within network footprint | Continental | Global | Regional–continental |
Best for | Asset tracking, fleet, GIS | Survey, construction, precision ag, autonomous robotics, drones | Fleet operations within network coverage | Automotive, fleets at scale, IoT | Remote/offshore where convergence delay is acceptable | Automotive ADAS, mass-market autonomy |
Point One product | DGNSS | Network RTK |
|
|
Choose RTK when you need centimeter accuracy with immediate convergence and operate within an RTK network’s coverage area. Surveying, construction machine control, autonomous vehicles, precision agriculture, and robotics are all typical RTK applications.
Choose PPP when operating in remote locations without RTK network coverage and you can tolerate a longer convergence period. Marine navigation, remote land surveying, and scientific applications in areas far from base stations benefit from PPP’s global availability.
Choose PPP-RTK / SSR when you need better-than-meter accuracy over wide areas with moderate convergence times. SSR’s separate error modeling makes it scalable for mass-market applications like automotive positioning.
For a detailed technical comparison, see our GNSS Correction Methods guide.
How to Choose an RTK Correction Service
RTK correction services support applications ranging from surveying and construction to autonomous vehicles, precision agriculture, and drone operations. The right service depends on your accuracy requirements, coverage needs, and operational environment. When evaluating providers, focus on these criteria:
Coverage and density. Verify that the provider has station coverage in your operating area — and that the station density is sufficient for consistent accuracy. Ask about future expansion plans if you anticipate operating in new regions.
Accuracy and convergence specifications. Compare horizontal accuracy (1–2 cm typical for quality services) and vertical accuracy (2–4 cm). Evaluate convergence time from cold start to Fixed — faster convergence means less downtime in the field.
Reliability and uptime. Production applications need 99.9%+ uptime. Ask about redundancy, failover systems, and how the provider handles station outages. Providers who own and maintain their stations typically deliver more consistent uptime than those relying on third-party infrastructure.
Hardware compatibility. Confirm the service supports RTCM 3.x and standard NTRIP protocols. Check that your specific receiver hardware is compatible and that mountpoint configuration is straightforward. For applications requiring positioning in challenging environments, evaluate whether the provider offers sensor fusion or INS integration.
Support and documentation. Evaluate technical support availability, developer documentation, API access for fleet management, and onboarding resources. For production deployments, responsive support can make the difference between a brief issue and extended downtime.
For a comprehensive provider comparison, see our NTRIP / RTK Service Provider Buyer’s Guide. For guidance on when to build your own base station versus subscribing to a service, see Build Your Own RTK. For drone-specific considerations including RTK vs. PPK workflows, see our Drone RTK guide.
Get Started with Centimeter-Accurate Positioning
RTK corrections are the technology that transforms standard GNSS from “good enough for navigation” into the centimeter-level precision that professional and industrial applications demand. Whether you’re guiding autonomous equipment, collecting survey-grade data, or building precision into your product, understanding how corrections work — and choosing the right correction service — is foundational to your success.
Point One’s RTK Network provides centimeter-accurate corrections across a dense, professionally managed network of base stations with 99.9% uptime. The platform supports single-baseline RTK, Network RTK (VRS), SSR, and DGNSS — all accessible through standard NTRIP with no proprietary hardware requirements.
Contact us at Point One sales to discuss enterprise deployments, coverage requirements, or custom integrations.
Frequently Asked Questions
What is an RTK correction service?
An RTK correction service is a subscription that provides access to real-time GNSS correction data from a managed network of base stations. Instead of building and maintaining your own reference station, you connect your RTK-enabled receiver to the service via NTRIP and receive corrections immediately. Services handle all infrastructure — station deployment, calibration, monitoring, and data delivery.
How accurate are RTK corrections compared to standard GPS?
Standard GNSS (GPS) delivers accuracy within 3–5 meters. RTK corrections improve this to 1–2 centimeters horizontally and 1–3 centimeters vertically — a roughly 100x improvement. This assumes RTK Fixed status, good sky visibility, and proximity to a base station within the network’s coverage area.
How far can I be from a base station and still get accurate RTK corrections?
For single-baseline RTK, accuracy degrades approximately 1–1.5 cm per 10 km, with an effective range of about 30 km. Network RTK (VRS) extends this significantly by generating virtual corrections near your location from multiple surrounding stations, maintaining consistent accuracy across the entire coverage area.
What equipment do I need to use an RTK correction service?
You need an RTK-capable GNSS receiver that supports RTCM 3.x corrections and NTRIP connectivity. Most modern survey-grade, machine-control, and robotics receivers meet this requirement. You also need an internet connection (cellular modem, WiFi, or ethernet) to receive corrections via NTRIP. For a list of compatible receivers and setup guides, see How to Choose the Best RTK.
What is the difference between RTK and PPP correction services?
RTK uses local base station infrastructure to deliver centimeter accuracy with convergence in seconds. PPP uses global satellite orbit and clock data — no local base stations needed — but requires 10–30 minutes to converge. RTK is better for applications requiring immediate precision; PPP is better for remote operations where RTK coverage is unavailable. PPP-RTK is an emerging hybrid that combines elements of both.
How much do RTK correction services cost?
Pricing varies by provider. Commercial RTK correction services typically range from approximately $50–$200 per month per device, depending on accuracy tier, geographic coverage, and volume commitments. Free alternatives exist (state DOT CORS networks, RTK2GO) but lack the reliability, coverage, and support needed for production applications.
Can I use RTK corrections without internet connectivity?
Yes, through alternative delivery methods. Radio modems can transmit corrections directly from a local base station (limited range, requires your own base). L-Band satellite delivery provides corrections via geostationary satellites without cellular or internet infrastructure — ideal for agriculture, maritime, and remote applications. For applications that don’t need real-time corrections, PPK workflows log data for post-processing.
What is the difference between RTK Float and RTK Fixed?
RTK Float is the intermediate state where corrections are being applied but the receiver hasn’t yet resolved integer ambiguity — accuracy is sub-meter but not centimeter-level. RTK Fixed means all ambiguities are resolved and the receiver is delivering full centimeter-level accuracy. Fixed is the target state for professional applications. Convergence from Float to Fixed typically takes seconds with a quality correction stream and good conditions.
What is the difference between OSR and SSR corrections?
OSR (Observation Space Representation) combines all error sources into composite corrections that the rover applies directly — this is the traditional RTK approach, compatible with virtually any RTK receiver. SSR (State Space Representation) models each error source separately and sends the components to the rover for reconstruction — this scales better for mass-market applications but requires specialized receiver firmware. Some modern services use SSR modeling internally but deliver corrections in OSR format for universal receiver compatibility.
Do RTK corrections work with all GNSS receivers?
RTK corrections require a dual-frequency, RTK-capable receiver that supports RTCM 3.x message formats. Most professional survey, machine control, robotics, and automotive-grade receivers meet this requirement. Basic consumer GPS receivers (smartphones, handheld units) generally cannot process RTK corrections. Check your receiver’s specifications for RTCM 3.x and NTRIP support.