A GNSS receivers’ function is to output a set of position coordinates after processing the reception of the signal broadcast by a constellation of GNSS satellites. Unofficially, the receiver’s grade used to be delineated by the accuracy that it could produce. A chipset would be capable of 5m CEP at best, while a geodetic class card could be tuned to achieve less than 1cm R95. With the advances in IC manufacturing and other innovations, however, that line is beginning to blur. In this blog post we compare these receiver grades along the signal processing chain, from reception at the antenna to a final position output.
Let’s begin with the antenna. In the geodetic class, the antenna element of the antenna connected to the receiver matters. This is the starting point where “garbage in – garbage out” is most applicable for RF designers. If you ever get the chance to look inside a geodetic class antenna, you may notice a whirlpool or vortex pattern in the element. While the fractal pattern is visually dazzling, the distance between the lines of the swirls is tuned and trialed painstakingly by RF engineers and designers to ensure the most stable phase center (the physical position in space at which the signals are received from the various azimuths and elevations of the satellites). This is the specific point of the coordinate. Manufacturers of such antennas are judged by the repeatability of the PCB routing with low tolerances.
As for the GNSS antenna in the mobile handset, its design is dictated by the amount of space available in the handset device, trading off phase center stability for meeting signal sensitivity requirements. Stable phase center variation is predominantly wishful in this use case due to radio wavelength physics. Furthermore, there are additional antennas in the handset that are connected to other transceivers that could interfere with GNSS reception, further fueling the chagrin of antenna designers already constrained by industrial design requirements.
Then there is the ability to track GNSS satellites. Correlation is the technique used to identify and lock onto a satellite. (Interestingly, it works very similarly to a song Identification service like Shazam and Soundhound.) In geodetic receivers, patented techniques are used singularly to lock onto satellites rejecting multipath signals to ensure a direct line of sight measurement. Consumer grade chipsets don’t have the computational resources to perform those advanced techniques and may even use a multipath signal. The shortcomings are mitigated with connectivity. Connecting to GNSS assistance services, like Location.io, to prime the receiver with information about the GNSS satellite’s location to increase its sensitivity in ranging to it.
The next part of the signal processing chain is measuring the distance to those satellites, correcting those measurements and, finally, determining an accurate location. In the geodetic case, typically two of the exact make and model GNSS receivers with very high measurement fidelity are paired together. One is configured as a reference base station receiver and situated at an optimal vantage point to observe as many satellites as possible and generate corrections for them. These corrections are then delivered via radio modem directly to the rover receiver. The rover applies corrections from the reference base receiver based on the common satellites between them and outputs an accurate corrected position.
The application of corrections also applies to consumer receivers. Both Geodetic and consumer grade rover GNSS receivers’ position accuracy benefits from corrections. But for consumer grade receiver use a typical user will not bother surveying the position of the base station reference receiver, or finding a survey monument to place the base station over, let alone configure the radio modem connection. Nor will they buy a reference base station receiver, nor check the matching make and model. The device just needs a source of corrections and, similarly to how GNSS assistance is applied, what the receiver lacks in measurements fidelity is made up for in connectivity. This time, though, they have access to reference networks that supply observations to generate corrections at a wider area scale. Corrections services like Truepoint.io are ready for this. A single base to single rover pair scales up to a reference receiver network serving many connected GNSS rover receivers.
At the top of this post, I mentioned the delineation between consumer grade and geodetic grade is getting blurred. Developments in reference network coverage, RF ASIC design, positioning algorithms, network bandwidth and connectivity, have all made it possible for accurate positioning to almost be taken for granted. GNSS chipsets now have comparable measurements to Geodetic grade receivers when connected to assistance and corrections services to deliver an accurate position to whichever app on the handset needs it. This is not to suggest that geodetic grade receivers are outdated. In fact, they play a vital role in reference networks with their configurability and dedicated computational horsepower for the different aspects of tracking a GNSS satellite to ensure integrity. The accuracy produced by consumer grade chipsets will always rely on the observations the geodetic grade receivers produce. Albeit a little blurry, the line delineating a geodetic and consumer grade GNSS receiver turns out to be a connection between them.