A simple comparison of various GNSS correction mechanisms
Let’s say you need reliable accurate global positioning in your technology. You do some research and decide to get yourself a multi-frequency GPS/GNSS receiver. You order an evaluation kit, but how to get your receiver to deliver the high accuracy that it promises? GNSS receivers rely on external corrections to compensate for various imperfections called GNSS errors to achieve decimeter or even centimeter level accuracy as fast as possible.
Correcting GNSS errors
GNSS based positioning is calculated using a method which by itself is limited in accuracy due to several errors caused by GNSS satellites as well as the Earth’s atmosphere.
GNSS satellites are essentially highly accurate synchronized clocks orbiting the Earth, constantly broadcasting their positioning and timing information. A GNSS user receiver gets signals from several of these “flying clocks” and calculates its distance to each satellite. When the receiver knows the distance to at least four satellites it can deduce its own position. However, the accuracy of this position is affected by certain errors.
Even advanced clocks on board GNSS satellites experience minute drifts which cause clock errors. The movement of GNSS satellites is predicted as they orbit the Earth. These predictions are not ideal, which results in what’s called orbit errors. Also, satellite equipment introduces small signal errors, which are modeled as satellite biases. In addition to satellite errors there are also atmospheric errors, which are caused by distortions and delays experienced by the signal as it passes through the Earth’s ionosphere (outer layer) and troposphere (layer near the Earth’s surface). Finally, the local environment around the receiver as well as the receiver itself can introduce errors. For example, satellite signals can be reflected off buildings and tall structures, a phenomenon referred to as multipath.
A GNSS receiver cannot correct satellite and atmospheric errors by itself and relies on data provided by an external source. Clock and orbit errors are satellite dependent, which means that they are the same around the world. Atmospheric errors, on the other hand, depend on the path the signal takes as it travels from the satellites to the user, and therefore differ depending on the receiver’s location.
To overcome both satellite and atmospheric errors a reference station also known as a base station, can be used. A reference station is a GNSS receiver installed at a fixed and precisely known location, estimating GNSS errors and sending them in the form of GNSS corrections to the user receiver (see image below). A reference network consists of interconnected reference receivers spread over an area.
Receiver-side errors can only be handled partially, by robust receiver technology and careful operation. Depending on which type of corrections are applied, it can take a few seconds to several minutes initialization time until high accuracy is achieved.
Types of corrections for high accuracy positioning
Until recent years RTK and PPP have been the established methods of providing GNSS corrections to user receivers. Nowadays the demand for high accuracy positioning is on the rise, paving the way for new positioning techniques such as the hybrid PPP-RTK.
RTK – the highest level of accuracy
In the RTK (Real Time Kinematic) method, a user receiver gets correction data from a single base station or from a local reference network. It then uses this data to eliminate most of the GNSS errors. RTK is based on the principle that the base station and the user receiver are located close together (maximum 40 km or 25 miles apart) and therefore “see” the same errors. For example, since the ionospheric delays are similar for both the user and the reference station, they can be cancelled out of the solution, allowing higher accuracy.
While in the RTK method corrections are provided for a specific location, in the PPP and PPP-RTK methods a correction model is broadcast to a larger area, but with slightly lower accuracy. To transmit this correction model a message format called SSR (Space State Representation) can be used. There is some confusion in the industry about the term “SSR” since it is often associated with the newer PPP-RTK method. But be careful, since “SSR” is occasionally used as a buzzword to refer to traditional PPP services as well.
PPP – globally accessible and accurate, but at a cost
The PPP (Precise Point Positioning) corrections contain only the satellite clock and orbit errors. Since these errors are satellite specific, and thus independent of the user’s location, only a limited number of reference stations is needed around the world. Since atmospheric errors are not included in PPP corrections, only a lower accuracy level can be achieved with this method, and a longer initialization time is expected, up to 20-30 min, which may not be practical for some applications. PPP has been traditionally used in the maritime industry and today has expanded to various land applications such as agriculture, as a convenient way to get global GNSS corrections.
PPP-RTK, the best of both worlds?
PPP-RTK (a.k.a. SSR) is the latest generation of GNSS correction services, combining near-RTK accuracy and quick initialization times with the broadcast nature of PPP. A reference network, with stations about every 150 km (100 miles), collects GNSS data and calculates both satellite and atmospheric correction models. As explained above, atmospheric corrections are regional, and so a denser reference network is needed than for PPP. These corrections are then broadcast to subscribers in the area via Internet, satellite or telecom services. Subscribed receivers use the broadcasted correction model to deduce their location-specific corrections, resulting in sub-decimeter accuracy.
Comparing the three GNSS correction methods
The table below compares the three correction methods, highlighting their strengths and weaknesses.
|Accuracy after initialization||~1 cm||2 - 8 cm||3 - 10 cm|
|Initialization time||Immediate||Fast (< 1 min)||Slow (~20 min)|
|Infrastructure density||~10 km||~100 km||~1000 km|
The infrastructure density and initialization time for all three methods vary with the different kinds of errors that are corrected, see image below. The broadcast nature of PPP-RTK and PPP, as well as the lighter infrastructure that they require, makes these methods scalable for mass market applications.
Some GNSS receivers also incorporate advanced positioning algorithms to compensate for receiver-side issues such as multipath (see Septentrio APME+), jamming and spoofing. This adds reliability and robustness to high accuracy positioning. For more information about spoofing see What is Spoofing and How to Ensure GPS Security.
Getting GNSS Corrections
Modern industrial receivers often get their GNSS corrections via a subscription service, delivered via Internet (using NTRIP protocol), satellite or 4G/5G. Today there is a boom in the correction service market driven by high accuracy demands of the automotive industry, automation and smart consumer devices. Automotive suppliers and many other new players are deploying infrastructure to set up services for centimeter-level positioning around the globe.
PPP and PPP-RTK corrections can even be transmitted directly by the GNSS satellites, as in the Japanese CLAS service from the QZSS constellation, or in the planned High-Accuracy Service (HAS) from Galileo. Depending on the network density and quality of the error modelling, different initialization times and accuracies can be achieved. This means that positioning quality can vary from one service provider to another.
Major telecom companies such as Deutsche Telekom as well as the Japanese Softbank and NTT are equipping their infrastructure with GNSS receivers to enable new corrections services. 3GPP, which provides specifications for mobile telephony including LTE, 4G and 5G, now covers broadcasting of GNSS satellite corrections in their mobile protocol. Since reference receivers are becoming part of critical infrastructure, such as telecom towers, it is essential that they have a high level of security to protect them from potential jamming or spoofing attacks (see Septentrio AIM+ technology).
Which corrections are right for me?
The right correction service for your technology will depend on your location and service area, your accuracy and reliability needs, as well as budget. Because the corrections market keeps expanding, it is now more important than ever that integrators or GNSS manufacturers assist you in selecting the best correction method for your industrial application. If you choose a GNSS receiver which does not “lock” you to a certain correction service, you will be free to choose a correction method which is most suitable for your application and its location. Such “non-locking” open-interface receivers also offer customers flexibility to switch to another more beneficial service in the future, as correction methods keep evolving.
 Global Navigation Satellite System including the American GPS, European Galileo, Russian GLONASS, and Chinese BeiDou, Japan’s QZSS and India’s NavIC. These satellite constellations broadcast positioning information to receivers which use it to calculate their location.
- PPP-RTK Technology Report, GSA
- Cooperation for future automated driving
- LTE Positioning and RTK: Precision down to the centimeter
CLAS - GNSS corrections for Japan - find out all about Japan's centimeter-level correction service
On-Demand Webinar: GNSS Corrections Demystified - get industry insights from guest speakers working at Fugro-Marinestar, Sapcorda, Lear Corporation and Point One Nav.