The microprocessor controls the entire receiver, managing its collection of data. It controls the digital circuits that in turn manage the tracking and measurements, extracts the ephemerides and other information from the Navigation message or CNAV, and mitigate multipath and noise, among other things.
The GPS receivers used in surveying often send these data to the storage unit. But, more and more, they are expected to process the ranging data, do datum conversion, and produce their final positions instantaneously, that is, in real-time. And then, serve up the position through the control and display unit (CDU). There is a two-way street between the microprocessor and the CDU; each can receive information from or send information to the other.
The microprocessor in a GPS receiver is the computer that manages data collection and is the home of the application of the software driving the receiver. The receiver also has storage, but, more and more, the microprocessor is expected to find the position of the GPS receiver in real time, instantaneously, or close to instantaneously, and serve it up in the control and display unit (CDU).
There are applications of GPS in which the receiver's microprocessor is expected to provide autonomous single-point positioning using unsmoothed code pseudoranges such as those from inexpensive handheld GPS units. Even though some manufacturers and users make extraordinary claims for their handheld C/A code pseudorange receivers, such autonomous point solutions are not accurate by surveying standards. But the positions from such receivers have improved since Selective Availability has been switched off. However, code-based pseudoranges using DGPS, differential GPS, can achieve good real-time, or post-processed results. DGPS is often used in collecting data for Geographical Information Systems, GIS.
Another method of getting instantaneous or close to instantaneous positioning from a GPS receiver is the navigation solution, as illustrated here, or the autonomous solution without differential positioning. I should mention another method that we'll talk about a bit later which is called Precise Point Positioning that will allow, using precise ephemerides, autonomous positions that are quite precise. However, it bears repeating that when an autonomous position is achieved with GPS receiver, it has to be based upon corrections provided in the navigation messages, which, as you know, have a certain amount of weakness.
The type of differential positioning sometimes known as DGPS depends on code pseudorange observations, but requires at least two receivers. One receiver is placed on a control station and another on an unknown position. They simultaneously track the same codes from the same satellites, and because many of the errors in the observations are common to both receivers, the errors are correlated and tend to cancel each other to some degree.
The data from such an arrangement is usually post-processed; although, with a radio link, results can be had in real-time. Improvements in this technology have refined the technique's accuracy markedly, and meter- or even submeter results are possible. Still, the positions are not as reliable as those achieved with the carrier phase observable.
With GIS, corner search and mapping work excepted, much GPS surveying requires a higher standard of accuracy. Certainly, GPS control surveying often employs several static receivers that simultaneously collect and store data from the same satellites for a period of time known as a session. After all the sessions for a day are completed, their data are usually downloaded in a general binary format to the hard disk of a PC for post-processing.
DGPS requires a base station at a known position, receiving signals from a constellation of satellites, the same constellation being received at the rover. That's important. A code phase solution requires at least four satellites, minimum. And there needs to be a link between the base and the rover to the correction signal. These signals can be received by the rover. The base station antenna needn't be on a building as illustrated. It could be on a tower. It could be a satellite. There are services that allow you to subscribe to a correction signal. In those cases, there is a network of base stations collecting the signals from the constellation; the correction signal is compiled, beamed up to the satellite, and that same message sent back down to subscribers from the satellite.
However, a DGPS solution need not be real time. In that case, of course, the base station could download to memory as does the the rover. Their data would then be brought together in a computer and post-processed.
Real-time kinematic surveying uses the carrier phase solution. Often, there is a radio link between the base and rover. While the baseline length with DGPS is often 100 to 200 kilometers or longer, the baseline length in RTK is more typically 10 to 20 kilometers. However, the arrangement of receivers is similar: the constellation of satellites being tracked at the base station and also tracked at the rover, the base station at a known point.
There is a transmission antenna used in RTK. It transmits the correction signals to the rover in real time correction. While the arrangement is similar to DGPS, the solution is carrier phase as opposed to pseudorange. The minimum number of satellites with RTK is five. It is five, so that one can be lost and still assured that there will be a solution.