October 16, 2003, 11:10 AM — Years ago, global positioning system (GPS) was an exotic service involving expensive equipment and used by the government and professionals in such areas as surveying, earth resources, and commercial transportation. And then, as I've mentioned before, along came the magic of VLSI, and the rest is, as they say, history. Inexpensive GPS chips (See example) have made it possible to put GPS into many consumer-grade devices, such as devices for tracking lost people, lost pets, a lost (or stolen) car, and providing one implementation of the E-911 service that allows mobile phone users to be located (much of the time, anyway) when they dial 911. GPS is practically a fixture in consumer electronics today, and used in many types of businesses on a truly global basis. Future location-based services will also be able to use GPS. Note that GPS isn't the only way to track people and things, but it's one of the most common, accurate and effective.
What is GPS? In a nutshell, it's a network of 24 "Navstar" satellites, orbited by the U.S. Department of Defense, and originally designed for, well, defense applications. GPS allows a user or device to locate itself on (or above) the surface of the earth with potentially very great accuracy (we'll return to the "potentially" in a moment). Each satellite is essentially an orbiting atomic clock (there are actually four atomic clocks on each satellite, for redundancy), and it broadcasts the value of this clock using a CDMA technique not unlike that used in CDMA cellular phones, but much more reliable. It also transmits several related pieces of data needed for accuracy on a continuous basis. On earth, a GPS receiver can detect this signal and, in fact, can often see up to 12 satellites at once, although only four are required for proper three-dimensional functionality with excellent accuracy. But the more satellites observed, the greater the accuracy in determining position. The calculation itself is quite simple - given the location of each satellite, all we have to do is measure the time delay between when the satellite sends a signal and when it is received on earth. Given that the speed of light is a constant, and the position of the satellites relative to one another is known, a triangulation calculation can then be performed, and voila, we know where we are.