RTK Float and Fix: Explained in 3 Minutes | Accuracy, Causes, and How to Tell Them Apart
By LRTK Team (Lefixea Inc.)
RTK positioning has two positioning states called the "Float solution" (浮動解) and the "Fix solution" (固定解). Can you correctly explain the difference between these two? The difference between them is a critical point that directly affects RTK positioning accuracy and reliability. In this article, we clearly explain in about three minutes the differences between the Float and Fix solutions in RTK, focusing on differences in accuracy, the causes of their occurrence, and how to distinguish them.
Table of Contents
• What is RTK?
• What is a Fixed solution (fixed solution)?
• What is a Float solution (float solution)?
• Difference in accuracy between Float and Fix solutions
• Causes of Float solutions
• How to distinguish Float and Fix solutions
• Easily achieve high-precision positioning with LRTK
What is RTK?
RTK (Real Time Kinematic) is a method that uses GNSS (Global Navigation Satellite System) to perform high-precision positioning in real time. Standalone GPS positioning typically has errors of several meters (several ft), but RTK uses two GNSS receivers—a base station and a rover—and the base station calculates positioning errors and sends correction information to the rover, enabling those errors to be eliminated down to the centimeter level (to within about 0.4 in). By installing the base station at a known, accurately surveyed coordinate and applying corrections to the satellite signals received by the rover at the observation point, RTK provides high-precision position coordinates in real time. This method cancels common error sources such as satellite orbit errors, clock errors, and atmospheric delays, typically reducing positioning discrepancies that were several meters to a few centimeters (a few in). Furthermore, high-precision positioning that uses carrier-phase measurements like RTK is far more accurate than conventional differential GPS (DGPS) based on code measurements (DGPS errors are at best on the order of a few tens of centimeters (a few tens of in)), and the fact that it can achieve centimeter-class positioning (about 0.4 in) is revolutionary.
To use RTK positioning, a reference station, a rover, and a communication link between them (radio or the Internet) are required, but if conditions are favorable, extremely precise positioning within a few centimeters (a few inches) is possible【for example, in environments where the reference station is nearby and satellite visibility is good, horizontal errors are about 1–2 cm (0.4–0.8 in) and vertical errors are on the order of several centimeters (a few inches)】. Because such centimeter-level accuracy (cm level accuracy (half-inch accuracy)) can be achieved, RTK is increasingly being used in a wide range of fields that require minimizing error, such as civil surveying, guidance of construction machinery, agriculture, autonomous driving, and drone positioning. In recent years, network-based RTK (such as the VRS method), which utilizes public and private reference station networks and allows correction information to be received without deploying one's own base station, has also become widespread. Specifically, methods such as VRS (Virtual Reference Station), which sets a virtual reference point near the rover via Internet distribution using cellular networks (Ntrip), are being provided in many areas. This makes centimeter-class positioning (cm level accuracy (half-inch accuracy)) possible without preparing a dedicated base station, and RTK is becoming an increasingly accessible technology.
What is a Fixed Solution (fixed solution)?
A fixed solution (fixed solution) in RTK positioning is the state in which the unknown integer-valued biases contained in the carrier phase of satellite signals (the so-called integer ambiguities) have been correctly resolved. Put simply, it is the stage at which the wave count of the radio wave included in the distance from the satellite to the receiver (an integer number of wavelengths) can be precisely identified; at this point RTK's inherent highest accuracy is realized. When a fixed solution is obtained, extremely precise positioning is possible: horizontal position errors are within a few centimeters (a few in), and vertical errors are on the order of a few centimeters to a dozen-some centimeters (a few in to a dozen-some in).
In RTK, immediately after positioning begins these integer ambiguities have not yet been fully resolved and the solution is unstable, but with current multi-GNSS-capable RTK systems it is common to reach a Fixed solution in tens of seconds to a few minutes under good conditions. By accumulating satellite data for a certain period and continuing the computations, the correct combination will eventually be found. Once the ambiguities are resolved as integers, that solution is confirmed as a "Fix", and thereafter high-precision positioning can be maintained unless there is a major signal outage. In many GNSS receivers and surveying apps, when RTK reaches a fixed solution the status display shows "FIXED" or "FIX", indicating that centimeter-level accuracy (half-inch accuracy) has been achieved. In operations using RTK, obtaining this Fix state becomes one of the objectives.
Note that older single-frequency (L1-only) GNSS receivers tend to take longer to reach a fixed solution and have difficulty maintaining a fix over long distances, but using modern multi-frequency, multi-GNSS-capable equipment can be expected to reduce initialization time and improve stability over long distances.
What is a Float solution (floating solution)?
A Float solution (floating solution) refers to an intermediate-stage solution in RTK positioning in which the integer biases have not yet been resolved. Although correction data are received and applied, the integer number of carrier cycles remains undetermined, so the accuracy is not fully stable. This Float solution is what is first obtained immediately after starting positioning or when satellite signal conditions are insufficient. Accuracy is improved over standalone positioning, but still remains roughly in the error range of ±0.5–1 m (±1.6–3.3 ft) (for reference, standalone positioning without corrections can reach errors of ± several meters or more). Compared with the centimeter-level Fix solution (cm-level accuracy, half-inch accuracy), larger deviations may occur — it is in the so-called “decimeter-class” accuracy (decimeter-level, about 3.9 in) and is not as reliable as a fixed solution, so as-is the Float solution is insufficient for precise surveying or machine control. Therefore, when using RTK, it is necessary to converge from the Float solution to the Fix solution as quickly as possible. However, in applications that can tolerate errors on the order of several tens of centimeters (several dozen centimeters, roughly 12–24 in) — for example, autonomous guidance in agriculture or general navigation — the Float solution may be sufficient in practice. Incidentally, the name “Float” comes from the fact that, unlike the solution fixed to integers (Fix), the solution is still computed as real numbers (floating point).
Differences in Accuracy Between Float and Fix Solutions
As mentioned earlier, there is a large gap in positioning accuracy between Fix and Float solutions. With a Fix solution, horizontal errors are confined to within a few centimeters (within a few inches), whereas with a Float solution errors of tens of centimeters to about 1 m (3.3 ft) can occur. This difference greatly affects the reliability of on-site positioning results. For example, in high-precision tasks such as establishing land boundaries or accurately locating structures, if positioning results are used while still at the Float solution, offsets of tens of centimeters (tens of inches) can occur and may lead to serious mistakes. In fact, there have been reports of surveys being carried out while RTK remained in the Float state, with large errors noticed later that required re-surveying.
On the other hand, in situations where accuracy requirements are not particularly strict, it may be more efficient to proceed using a Float solution obtained in a short time. However, even in such cases, you should be aware that the results obtained can contain relatively large errors. Basically, to obtain accurate results with RTK, it is safer to wait until it switches from Float to Fix before using the positioning results. For positioning that requires high precision, you should adopt the results only after confirming that the solution is in the Fixed state.
Therefore, when performing RTK positioning, it is extremely important to confirm that the solution is in the Fixed state before adopting the results. Only when a Fixed solution is obtained can RTK's centimeter-level accuracy be guaranteed, allowing you to use the positioning results with confidence. Conversely, while the solution remains in the Float state, the accuracy is still unstable and preliminary, so you must wait patiently for the positioning engine to converge to a Fixed solution. Even after achieving a Fixed solution, the result can revert to a Float solution due to various factors; in that case, remain calm, identify the cause, and wait for the solution to become Fixed again before resuming work. Being able to distinguish between Float and Fixed states and respond appropriately is a fundamental principle for operating RTK correctly.
Causes of Floating-Point Results
Normally, if the sky is clear and a sufficient number of satellites are visible, the solution will enter a Fix state within tens of seconds to a few minutes after starting RTK. If it does not reach Fix after more than 5 minutes, it is highly likely that some factor is preventing the solution from converging. When RTK positioning remains in a Float solution and never fixes, the main typical causes to consider are as follows:
• Poor satellite reception environment: if the sky view above is restricted or there are surrounding buildings or trees, satellite signals can be blocked and the number of receivable satellites becomes insufficient. Also, if reflected radio waves from walls or the ground (multipath) are picked up, errors increase and the RTK engine may not be able to converge the integer bias, leaving the solution in a Float state.
• Errors in reference station installation or coordinates: if the correction information received by the rover itself contains errors, it may not be possible to reach a correct Fixed solution. Installing the reference station on a point that is not an accurately known point, or having an incorrect initial coordinate setting for the reference station, can cause the solution to be unstable and prevent it from becoming Fix.
• Reference station distance too long: if you are too far from the reference station that provides the correction information, the error sources contained in the satellite signals received by the reference and rover will differ significantly, reducing the effectiveness of the corrections. In general, for RTK using a single reference station, a distance (baseline length) to the reference station of about 20 km (65,616.8 ft) or less is desirable, and if it is too far it becomes difficult to converge from a Float solution to a Fix solution. Note that high-performance receivers may be able to achieve Fix even at distances beyond 70 km (229,658.8 ft), but the larger the distance the greater the error and instability of the solution; for example, there are reports that a baseline of 100 km (328,084.0 ft) can produce a positional shift of about 10 cm (3.9 in).
• GNSS setting mismatch: if the satellite systems (GPS, GLONASS, Galileo, QZSS (Michibiki), etc.) or frequency bands used by the reference and rover are not the same, the rover may not have the necessary observation data and integer ambiguities cannot be resolved. (For example, if the reference station is distributing corrections using GPS and GLONASS but the rover is configured to use only GPS satellites, the necessary observation data will be missing.) Using the same satellite and frequency settings as the reference station is a prerequisite for acquiring a Fix.
• Correction data not received / communication trouble: if RTK correction data from the reference station is not being received, you cannot advance beyond a Float solution no matter how good the observation environment is. If radio or internet communications are unstable and corrections are missing or interrupted, the RTK engine cannot hold the solution and the floating state will persist without reaching Fix.
• Other factors: misconfiguration or malfunctions of GNSS equipment or software can also prevent obtaining a Fix solution. For example, if the rover’s mode is not set to RTK (differential positioning), if the communication format between the reference and rover does not match, or if there is equipment trouble such as a broken antenna cable, you may be unable to progress beyond a Float solution.
If you suspect any of the causes described above, taking appropriate measures—such as improving the on-site reception environment or reviewing device settings—can increase the likelihood of achieving a fix.
How to distinguish Float and Fix solutions
So how can you tell in the field whether your positioning solution is Float or Fix? The most reliable way is to check the status display of the GNSS receiver or the application you are using. Many devices and software explicitly indicate the RTK solution status as "FIX" or "FLOAT" (when no corrections are available they may display "Single" or "DGPS", etc.). If the screen shows "FIXED" or "fixed" it is a Fix solution; if it shows "FLOAT" or "float" it is still a Float solution. Depending on the model, the status may be shown by indicator colors (e.g., green for Fix, yellow for Float) or by numeric quality indicators, but basically you can easily distinguish the state from the display. (By the way, inside the RTK engine a solution is judged to be Fix when a value called the "AR ratio" exceeds a certain threshold, but for typical users checking the device's status display is sufficient.)
Also, you can roughly distinguish by looking at the estimated positioning accuracy (position accuracy display). For example, if the current accuracy shown by an RTK receiver or an app is at the level of horizontal ±0.02 m (±0.07 ft), it can be considered in a Fix state, but if it shows a relatively large error such as ±0.5 m (±1.6 ft), it is likely in a Float state. Furthermore, in a fixed solution the variation of positioning values when stationary is contained within a few centimeters (a few inches), whereas in a float solution the position tends to drift on the order of several tens of centimeters (several tens of inches). In any case, for positioning that requires high accuracy, always confirm before starting work that the status is Fix, and if it reverts to Float during operations check the cause and wait for it to return to Fix before resuming. Correctly reading the device display and accuracy information to distinguish Float/Fix is a practical basic.
Easily achieve high-precision positioning with LRTK
In recent years, devices have emerged that allow the use of RTK positioning without specialized knowledge or complicated setup. A typical example is the small GNSS RTK terminal "LRTK", which can be attached to and used with an iPhone. LRTK is a system that links with a smartphone to enable one-touch centimeter-level positioning, automatically performing high-precision correction processing internally. By attaching the dedicated receiver to an iPhone, users can start high-precision positioning on site immediately without having to worry about installing a base station or configuring complex communications.
For example, if you attach an LRTK device to the supplied lightweight pole (monopod) and survey, a single person can efficiently observe many points in a short time. Also, because the smartphone app automates the antenna height (the height at which the device is placed) correction calculation, you can obtain accurate ground-surface coordinates without specialized knowledge. Furthermore, the LRTK device itself has a built-in battery and is so compact and lightweight that it fits in the palm of your hand, making it highly portable and easy to handle even in confined sites.
Specifically, the LRTK terminal automatically retrieves correction information, such as the Geospatial Information Authority of Japan’s electronic reference point data, over the Internet via its built-in communication module, and by simply pressing the "Start Positioning" button on the smartphone app you can obtain an RTK fix. Under favorable conditions, high-precision positioning is possible, with horizontal position errors on the order of ±1–2 cm (±0.4–0.8 in) and vertical errors on the order of ±3 cm (±1.2 in), enabling you to easily achieve accuracy comparable to conventional professional GNSS equipment. RTK technology is no longer limited to a handful of specialists; tools like LRTK are making it possible for anyone to benefit. If you are thinking "I want to introduce RTK into my own work" or "I want to improve positioning accuracy to increase work efficiency," you might consider adopting such easy-to-use RTK solutions.
Please correctly understand the difference between RTK Float solutions and Fix solutions, and use that knowledge to help achieve high-precision positioning in the field.
Next Steps:
Explore LRTK Products & Workflows
LRTK helps professionals capture absolute coordinates, create georeferenced point clouds, and streamline surveying and construction workflows. Explore the products below, or contact us for a demo, pricing, or implementation support.
LRTK supercharges field accuracy and efficiency
The LRTK series delivers high-precision GNSS positioning for construction, civil engineering, and surveying, enabling significant reductions in work time and major gains in productivity. It makes it easy to handle everything from design surveys and point-cloud scanning to AR, 3D construction, as-built management, and infrastructure inspection.


