How much is 15.6 centimeters worth? That’s the width of U.S. paper currency, which is worth anywhere from $1 to $100. But in the case of construction, agriculture, logistics, and other verticals, 15.6 centimeters can easily be worth thousands of dollars — or more.
That’s why dozers and other heavy construction equipment now use GNSS to tell operators when they’ve achieved exactly the right grade, thus eliminating wasted fuel, labor, and materials applying more than the job specs require. On farms, drones and planters use GNSS to put the precise amounts of seed, fertilizer, and pesticides in each row. And in ports and logistics parks, GNSS helps ensure that autonomous material handlers always have the right intermodal container in the right place.
But they’re not using standard GNSS data, which is accurate to only about 0.5 meter. Instead, they’re augmenting that data with a correction service to achieve accuracy of 15.6 cm — or even higher.
Correction services are necessary because a host of factors undermine GNSS accuracy. For example, the signals are weak by the time they reach the Earth, so they’re vulnerable to atmospheric interference. Two other challenges are satellite clock/orbit errors and multipath effects.
Read on to learn about the major types of GNSS correction services, including how they work and how to add support for them in your IoT devices.
The Correct Answer
When the world’s first GNSS constellation — GPS — was opened to non-military applications in 1983, civilian receivers were limited to an accuracy of about 100 meters (330 feet). In 2000, that restriction was eliminated.
Another improvement was additional signals. GPS initially had just one signal for civilian applications: L1 at 1575.42 MHz, which enabled location accuracy of 3 meters. Today, L1 can be combined with the newer L2 (1227.60 MHz) and/or L5 (1176.45 MHz) signals to increase accuracy down to 1.5 meters. That’s because L2 mitigates ionosphere-related errors by providing a second signal to augment L1. The L5 signals also are more powerful than L2 signals. (For a deeper dive, see “Navigating the L1, L2, and L5 Band Options for GNSS”.)
| Item | Description | Typ. Error |
|---|---|---|
| Satellite Clock Errors | Position depends on clocks. Each satellite’s clock can wander. Correction information is sent down from each satellite. Without correction, this error can be up to 300 km. | 0.4 – 1 m |
| Satellite Position Errors | Position also depends on knowing the position of each satellite. The satellites transmit their own position (ephemeris) but this isn’t perfect. | 0.3 – 1 m |
| Ionospheric Delay | The upper layers of the atmosphere are “ionized” by the sun, which interacts with signals sent between satellites and the Earth. Stand-alone receivers can use a mathematical model to provide some correction. Without correction, this can be 7 m. | 1 – 3 m |
| Tropospheric Delay | The Troposphere is the lowest layer of the atmosphere and where we live. Rain, fog, and other water in the air delays the signal. This delay varies by location, height, and angle to the satellite. Models can be used to reduce the error. | 0.2 m |
| Receiver & Antenna Biases | Receivers have biases that introduce errors. These are typically small, on the order of cm. Antennas can also introduce biases (phase center and group delay). | 0.2 m |
| Multipath | As signals travel from the satellite to the Earth, they bounce, reflect, and distort. | 0.2 m |
| Total | Single-Frequency Receiver | 2.3 ~ 5.6 m |
| Dual-Frequency Receiver | 1.5 ~ 2.8 m |
But even 1.5 meters isn’t granular enough for autonomous trucks, robotaxis, surveying, and other demanding business applications. That’s where correction services come in. As their name implies, these mitigate a variety of factors that undermine GNSS accuracy, such as signal attenuation due to dense foliage and tall buildings, atmospheric interference, and clock orbit errors.
One option is an L-Band correction service, whose data augments GPS and other GNSS constellation signals to achieve accuracy of about 20 centimeters. L-Band services are often referred to as precise point positioning (PPP) services. (For more information, see “How to Leverage the L-Band to Balance Accuracy and Affordability for GNSS Applications.”)
If the application requires accuracy as low as 1 centimeter, the best solution is Differential GNSS (DGNSS), where terrestrial reference stations broadcast location data. One example is real-time kinematic (RTK), a subscription-based correction service. PPP and RTK also can be combined in a hybrid approach. (For more information, watch the on-demand webinar “High Precision GNSS and RTK Positioning.”)
PPP and RTK services require a subscription, which can be an issue for cost-sensitive applications. One free alternative is the Galileo High Accuracy Service (HAS), which provides real-time orbit, clock, and bias data to achieve up to decimeter-level accuracy. The first phase of HAS launched in 2023 and delivers correction data via satellite broadcasts and the internet. (For more information, see this FAQ from the EUSPA.)
One drawback to satellite-delivered correction services — both free and paid — is that their signals are susceptible to the same vulnerabilities as standard GNSS. For example, if dense foliage and concrete canyons severely attenuate the base GNSS signals, then it’s likely that the correction signals also will suffer. Another potential risk is jamming and spoofing of the correction signals, which is increasingly common with base GNSS. (For more information, see “Top GNSS Jamming and Spoofing Attacks and How to Mitigate Them.”)
The Role of Antennas
The right antenna is critical for wringing maximum accuracy out of both raw and corrected GNSS data. But “right” is relative.
For example, in vehicular applications, the antenna likely will be able to use the metal roof or trunk lid for its ground plane, making patches and crossed dipoles two good choices. An example of the latter is the Taoglas EAHP.50.
In other applications, no ground plane will be available, which means a quad helix antenna is the way to go. Two examples are the Taoglas EAHP.125 and the Taoglas Accura TS.125.
If it’s difficult or impossible to ensure that the antenna will always face up at the sky for a clear view of multiple satellites, an omnidirectional model is the way to go because it’s designed to pull signals from any direction.
For more insights into how to choose the right antenna, and then integrate it to maximize performance, check out the following blog posts:
- “Six Key Parameters to Consider When Comparing GNSS Antennas”
- “GNSS Antennas: How to Choose Between Directional and Omnidirectional Designs”
- “Maximizing GNSS Antenna Performance: Integration Tips for Engineers”
- “The Right GNSS Antenna is Key for Balancing Low Power and High Performance”
- “Four Top Reasons Your GNSS and Cellular Antenna Performance is Suffering”
Next, use this tool to see which Taoglas GNSS antennas meet your requirements.
Engineering expertise also is critical for integration, testing, and optimization. That’s why Taoglas offers engineering services such as:
