There is no shortage of information on how GPS works. My explanation is meant as a brief summary. If you would like more information, I have many sources in the links and reference section.

Essentially GPS takes the range (distance) from a constellation of satellites to calculate your position. If the GPS knows where you are and you tell it where you want to go it is then able to calculate bearings, distances, off course distances, and your track.

If you want to have a slightly more detailed explanation, I have a slightly longer but still purposefully over simplified explanation in the next several pages. If you want a more detailed explanation than I offer, here are some sources:

• GPS Explained, Paul Bertorelli of IFR magazine http://www.avweb.com/articles/gpsexpln.html One of the best explanations that I have seen on how GPS works. • GPS Guide for Beginners, http://www.garmin.com/aboutGPS/manual.html This is a Garmin pamphlet that is good at explaining the basics of how GPS works. There are also basic introductory explanations of elementary navigation terms. • Trimble GPS tutorial, http://www.trimble.com/gps/index.htm This is a Flash explanation of how GPS works. • Navtech GPS Seminars and Supply, http://www.navtechgps.com If you want a post graduate engineering text on GPS, Navtech would be a source. Basic GPS

The GPS system consists of 24 satellites. The number may vary slightly as new ones are launched and old ones are retired. Each satellite is in an 11,000 mile orbit and transmits a very weak signal. The system is monitored and maintained by the U.S. Military. The satellites only broadcast to the user and the user only receives. There is no charge for use.

To start with, assume that all of the satellites and the receiver have a perfect internal clock. This is not the case, but it makes a good starting point. Each satellite transmits a coded signal. Consider this signal to be like the peaks and ridges along the edge of a super long key. This code is generated as a function of time. The receiver is also able to generate the same code. The receiver matches the incoming code to the internally generated code except that there is a delay caused by the signal’s travel time between the satellite and the receiver. The receiver measures how much it has had to shift the timing of its code to match the incoming code. Since the receiver knows how much time it took the signal to reach the receiver and the speed of travel of the signal, it can then calculate the distance from the satellite.

If you know how far you are from one satellite then you know that you are somewhere along an imaginary sphere around that satellite. If you know how far you are from two satellites, then you are somewhere along the intersection of where these two spheres, which is a circle. If you add another satellite, then you are somewhere where this third sphere intercepts the circle created by the intersection of the other two spheres. The sphere will most likely intercept the previous circle at two points. One of these points is where you are, and the other is not a reasonable solution – somewhere in outer space. Thus by knowing where you are relative to these three satellites the receiver with a perfect clock can know where it is.

Although no clock is perfect, the satellites have atomic clocks—pretty close. The clock in the GPS receiver is closer in technology to an inexpensive digital watch. Light travels at 186,000 miles per second. If the receiver time was off by 1/100 of a second the calculated distance would be off by 1,860 miles.

For each receiver to have its own cesium clock would make GPS technology prohibitively expensive and non-portable. What the GPS receiver does is to use a cheap clock similar to a digital watch and add one more satellite to the calculation to correct the time in the receiver. The receiver shifts the time calculation back and forth so that all of the imaginary spheres around the satellites intercept at one point.

For three-dimensional navigation you need to receive four satellites. Think of it as one satellite for each dimension and one for the time. For two-dimensional navigation you can scrape by with only receiving three satellites. If you know your altitude, the GPS can treat the center of the earth as a satellite reducing the number of required satellites by one. Your distance from the center of the earth is the radius of the earth plus your altitude. This is why aviation GPS models have barometric altimeter input and you may occasionally see a handheld GPS ask for your altitude during poor reception conditions.

Newer GPS receivers use the extra signals above the minimum that is required to further refine the position for increased accuracy. This is known as an over determined solution.

Selective Availability, SA

SA is an intentional error introduced into the GPS signal to make it less accurate. Although I suppose that the military could turn it on again, SA no longer exists. I mention it because you may see it in mentioned in literature on GPS.

Not only is GPS good for flying airplanes, but it is good for guiding bombs and missiles. To prevent somebody else from doing this well, the military added a little random time shift to the satellite signal available for civilian use. This added some inaccuracy to the calculated position.

Error correction technologies such as differential GPS, WAAS, and LAAS take out much of the SA induced error. Thus a sophisticated enemy could negate the effects. Thus, selective availability was turned off.

Differential GPS

If you have an inaccurate piece of equipment, but know exactly how inaccurate the output is, then you know the correct value. For example, if your watch was exactly 5 minutes fast, you could look at your watch and subtract five minutes to know exactly what time it was. In fact, I have found most people who use this watch setting technique to prevent chronic tardiness also subconsciously perform this calculation every time they look at their watch.

Differential GPS technologies use a similar idea. The signal from each satellite must pass through the atmosphere. The atmosphere, and especially the ionosphere, causes errors due to refraction. The GPS receiver has some internal models to calculate these effects, but an even better way is to directly measure the errors.

The idea of differential GPS is to install a GPS receiver at a known point. Since the location is known, this GPS compares the distance to each satellite and to what it should be and then rebroadcasts the error in digital format. GPS receivers so equipped can then use this known error in its position calculations.

Basic GPS

Basic GPS uses local receivers and local transmitters. In order to use differential GPS, you need a special differential receiver which then sends the signal to the the GPS. Differential GPS is mainly a marine application and it is not widely used for recreational applications.

For more information on differential GPS:

http://www.navcen.uscg.gov/dgps/default.htm

One of the biggest advantages of differential GPS is that it helped to eliminate the purposeful errors caused by Selective Availability. Now that Selective Availability has been turned off, the level of accuracy increase from using differential GPS is significantly less. Additionally, WAAS, Wide Area Augmentation Service is common in most new GPS receivers. WAAS is a differential type of technology.

WAAS

Another differential technique is known as WAAS, Wide Area Augmentation Service. WAAS has 25 receivers scattered around the United States. A mathematical model of the satellite errors is created based on the measurements and the error correction values are then sent to a geo synchronous satellite to be rebroadcast. The advantage of WAAS over conventional differential GPS is that it is available in small handheld receivers without needing a separate receiver. In fact, almost every new GPS receiver is WAAS capable.

Europe is developing a system similar to WAAS called EGNOS, European Geostationary Overlay Service. Japans is developing MSAS, Multi-Function Satellite Augmentation System. Hopefully, any WAAS receiver should work with EGNOS or MSAS.

WAAS was designed for aviation use. GPS is more accurate in laterally than for altitude. One of the goals of WAAS was to provide sufficient accuracy to allow GPS to be used to provide vertical guidance during an instrument approach. Most larger airports have something called an ILS (Instrument Landing System) which provides a “radio beam” down to the runway. However, many smaller airports do not have this expensive navigational infrastructure. There are many issues involved, but WAAS enabled GPS approaches with vertical guidance offer big safety improvements to runways without ILS equipment.

The other goal of WAAS is not so much accuracy as it is integrity. If a satellite is sending a bad signal, it takes a few minutes to detect and stop broadcasting the signal. Currently aviation receivers use satellite signals beyond the minimum required to cross check the accuracy of the signal. For example, if you need 4 satellites to determine a position and you are receiving 5 satellites, you can use the extra signal as a cross check. This is called RAIM, Receiver Autonomous Integrity Monitoring. Part of WAAS provides integrity checking which is faster than what is offered through the basic GPS system.

Most newer inexpensive handheld GPS receivers, use the extra signals above the minimum that is required to further refine the accuracy of the position solution. The technical difference between RAIM and the possibly proprietary algorithms that consumer handhelds use is well beyond the scope of this discussion or my knowledge. However, I think that it is fair to say that an aviation receiver is optimized to give as quick of a warning as possible to bad or insufficient satellite data, whereas a consumer GPS and the aviation handheld receivers that they are based on them are not designed with this in mind. In fact, consumer handheld receivers are probably more designed to not give nuisance warnings than they are to give timely warnings of bad navigational data. This is not necessarily bad design as much as it is a reflection of differing design parameters for different uses. For the most part, this is not a big issue, but it is a very good reason why you cannot use a handheld receiver as if it were a certified aviation receiver.

What your GPS does when it starts up

You may have noticed that the amount of time it takes for your GPS to calculate a position varies. It will take an especially long time to get an initial fix when you first start it and it will get a fix very quickly when you start it again after just shutting it down.

The GPS has two types of data on the location of the satellites and their orbits. The first is a rough idea of where each satellite is located and is called the almanac. This almanac data is good for a couple of months. If the GPS does not have a current almanac it will take about 15 minutes to download.

The second type of data is the fine data more technically referred to as the ephemeris data. Each satellite broadcasts the almanac which is applicable to all of the satellites, but only broadcasts its own ephemeris data. The ephemeris data takes 18 seconds to download and is good for a couple of hours. It is this ephemeris data that the GPS actually uses for deriving a position. The almanac is used for deciding which satellites to “look for.”

For most 12 channel parallel receivers, the GPS will start looking for the satellites that it expects that it can receive based on it’s current position and time using the almanac data. The GPS assumes that it is where it was last shut down and the clock is correct. However, you can change the position and time, this is called initialization.

No accuracy is required in this initial position. I have shut down my GPS in Florida and turned it on in Europe and was able to get a position. The GPS did not attempt to look for satellites that would be in view to the east because they would be invisible from Florida where the GPS was assuming that it was. Likewise, the GPS was attempting to download data from satellites that would be well over the western horizon from Europe that would be visible from Florida. However, there are usually enough satellites that would be visible from both Europe and Florida and eventually the GPS will get a position and sort things out.

I have been keep the GPS from getting a lock in the wide open outdoors by initializing it to the other side of the world. Thus none of the satellites that it was attempting to receive would be in view. The point of this is that when you give the initial position during the initialization, accuracy is not important – anywhere within a couple of thousand miles is probably good enough.

Obviously, part of this calculation is the almanac data. If the Almanac data is grossly out of date, the GPS will not have the correct data to calculate which satellites to look for. The almanac data takes 12.5 minutes to download. Thus, you should leave the receiver on for at least 15 minutes to a half hour every couple of months to get a fresh almanac.

Most receivers have a mode where they can just start searching cycling through the list of satellites searching in a trial and error manner. The advantage of this mode is that it does not depend on a initialization position, time, or current almanac.

Most receivers will display a list of choices if it has trouble getting an initialization asking you if you want to use the automatic mode, enter a new position, continue trying with the same initialization, or just give up because you are indoors.

Having more than 12-channels will make most of this discussion a non-issue. Cobra makes a handheld GPS with 18 channels at the time of this writing. I would be surprised if other manufactures don’t eventually follow. Quite honestly, getting an initial first fix usually is not a problem. When it is a problem, it is simply dealt with by giving the GPS a new position or using the automatic mode.

Once the GPS starts to receive data from a satellite, it will show a hollow bar on the satellite page. On some receivers you might see bars go solid with others following. On other receivers you might not see any go solid until at least three go solid simultaneously. In the first case, the bars go solid as soon as the ephemeris data can be used to give a pseudorange to calculate a position. In the second case, the solid bar indicates that the satellite is being used for a position fix. In this second case, If at least three satellites are not being received with current ephemeris data, there is no position fix and therefore an individual satellite is not being used for a position fix because there is none.

A “D” for differential superimposed on the bar means that WAAS corrections are being applied.