|
GPS Consumer Series: Selecting the Best Type of GPS Data for you Application
City managers in Fort Collins, Colo. integrate their GIS with digital orthophotos for positive results.
By Chuck Gilbert Introduction
In the minds of many novice GPS users there is only one kind of GPS position. If a position was computed by a GPS receiver then it's a GPS position and that's that. In reality, there are several ways that a position could be generated from GPS satellite data. This month, this column examines two types of GPS data and the difference between the positions that are computed from each. Two choices
The data from GPS satellites contain many types of information that can be recorded. Some types of GPS satellite information include multiple codes generated by each satellite, and information about the radio signal transmitted from the satellite (such as Doppler shift and signal strength). Most GPS receivers do not store every bit of the available information that comes down from the GPS satellites. Although they all ultimately compute a position, most are limited to storing/using only certain types of information to compute that position.
GPS data can be grouped into two broad categories; code data and carrier data. Code data are computed from a code that is generated by the GPS satellite, then transmitted to the user on a radio signal. Carrier data are computed from examining the actual radio signal that was used to carry information from the satellite to the user.
Some simple receivers are locked into storing only code data, other more sophisticated receivers give the user a choice about whether to store carrier data or code data. Both of these data types can ultimately be used to compute a position, but they are used differently and can impact your work flow in several respects. These differences are discussed in the following sections. It takes two to Tango
Let's consider for a moment the difference between positions, and accurate positions. In either case, recording code or carrier, data from a minimum of two GPS receivers are required to compute accurate positions. With only one receiver, you cannot compute a carrier position, however you can compute a code-based position that is accurate to about 100 meters. This is the specification of Selective Availability (S/A). S/A is the intentional degradation of the GPS signal to "less than 100 meters 95 percent of the time" by the U. S. government. To achieve accuracy significantly better than 100 meters, all GPS receivers require a second set of GPS data collected by another receiver that was stationary at a location of known coordinates. Such a receiver is often referred to as a base station.
If the roving receiver recorded code data, the base and rover data can be processed together to improve the rover accuracy from 100 meters to a few tens of centimeters (known as differential processing). After such processing data from a high quality receiver will result in code positions as accurate as 30 - 40 centimeters. A lower quality code receiver will result in code positions with an accuracy ranging from 2 - 5 meters after differential processing.
Alternatively, if the roving receiver recorded carrier data, the data from both receivers could be processed together to create positions accurate to better than 1 centimeter! Positions computed from carrier processing techniques can be tremendously accurate. The bottom line is an application that requires accuracy consistently better than about 30 centimeters, requires carrier-based GPS data. An application that can accept accuracy from 0.3 to 5 meters requires code-based GPS data. How far can I go?
Why, you might ask, would anybody consider a code-based GPS receiver if carrier data are so much more accurate? One answer lies in a constraint that come along with differential GPS data collection. When using data from two receivers, the distance between those receivers can become an important factor. Most users do not consider the use of carrier processing if the base/rover distance is greater than about 50 kilometers. At distances greater than a few tens of kilometers carrier processing may be unreliable.
Additionally, the error of the processed GPS data will increase as the distance between the base and rover receiver increases. The magnitude of this degradation is typically on the order of a few parts per million (ppm). For example, if the precision degrades at 2 ppm, then at a distance of 50 kilometers a user could expect an additional 10 centimeters of error (50 km x 2 ppm). The user who employs carrier techniques in the hope of achieving accuracy of only a few centimeters will often find that the advantage of carrier processing may be significantly eroded at distances of 50 kilometers. The combination of these factors results in people typically using carrier techniques only within about a 50 kilometer radius of their base station.
Code techniques, on the other hand, are often considered usable to distances of over 1,000 kilometers from the base. This allows one base station to provide vast areas with reasonably accurate code-based positions. The same magnitude of degradation still applies, for example 2 ppm. However, a user who aspires to a code accuracy of 1 meter, at 300 kilometers from the base, might still expect sub-meter results since the additional error (due to base/rover distance) is only about .6 meters (300 km x 2 ppm). Can I work here?
The work environment is another important consideration in deciding whether code or carrier is more appropriate for your application.
Successful use of GPS carrier data is contingent upon continuously tracking as many satellites as possible for as long a time as possible. Carrier data are best collected in an area that has an open view of the sky. An area with a low horizon in all directions and few sky obstructions is ideal.
This issue is one of the fundamental differences between carrier data collection and code data collection. In order to compute an accurate position from the GPS code the receiver need only track four satellites for some fraction of a second. (To be conservative, let's just say one second.) This is true for even a very low quality code-based receiver.
Carrier-based systems vary widely on this point. A low end carrier-based receiver may require tracking several satellites (the more the better) continuously for several minutes to collect enough data to compute a single position. A high end carrier-based system could compute a position in one second, but only after the receiver has first been able to become initialized by tracking several satellites continuously for about a minute or more.
In the scenarios described above, pay particular attention to the use of the word "continuously" when describing the satellite tracking. This is perhaps the most important point to be made here. If the data from a satellite are being periodically interrupted by nearby obstructions, that satellite is probably not usable for carrier processing. Therefore, there are some real world environments where doing carrier-base GPS data collection is simply not an option.
For example, in a forest, the trees will generally disrupt the GPS signal to such a degree that carrier data collection is not feasible. However, code positions can be collected under most tree canopy situations; the accuracy will be slightly degraded, but positioning is possible.
Another example would be using GPS to record a major highway. Most users who want to record a highway would do so by simply driving the road with a GPS receiver attached to the vehicle. The problem is that in the process of driving a major road you may frequently pass under bridges. Every time you pass under a bridge the GPS satellite signals are momentarily interrupted. For a code-based receiver this would not be a problem. There might be a small gap where there are no positions computed directly under the bridge. However in many cases this may not even be detectable. The positions immediately before and after the bridge would all be of equal and acceptable accuracy.
On the other hand, with a carrier-based receiver, the accuracy of the positions would be a direct function of how long you could collect continuous satellite data (e.g. how much time between bridges). If, on average, you pass under a bridge every 2 minutes you will probably not get any useful carrier data unless you have a very expensive high end carrier-based system. Using a low end carrier-based system (closer in price to a code-based system) you would require some tens of minutes of data between each bridge in order to achieve decimeter accuracy. GPS has proven to be a very powerful tool for mapping roads. However, for the reason described above, most dynamic road mapping is done with meter level, code based systems. Conclusion
Carrier-based GPS positioning is a powerful tool. The accuracy of carrier-based positioning is excellent, however there is a price. Next issue's column will continue this comparison of code positioning versus carrier positioning. Specifically - addressing how these comparative strengths and limitations impact the planning and execution of field work.
Back
|
