Feb. 8, 2006
By Josh Blair
If someone knew they would be wrong only one time out of a billion, they might want to consider moving to Vegas.
Yet, for an aircraft to land using Global Positioning System (GPS), it must be proven that the GPS signal is correct 999,999,999 times out of a billion. Yes, that is nine nines, and as you can imagine, proving something that many times is not an easy thing to do.
GPS is a satellite navigation system used for determining precise locations that consists of a constellation of 24 satellites that circle the Earth twice a day. Obstructions in the air and on the ground can cause the GPS signal to be faulty. But thanks to senior electrical engineering major Clint Barker and his work with research engineer Curtis Cohenour, this task is going to be a little bit easier.
The two are using a self-constructed hemicap, which is an eight-foot-long, four-feet-in-diameter, Tylenol-shaped capsule made out of sheet metal, to verify their GPS multipath model. Multipath is the phenomenon where the GPS signal at the receiver is a combination of the direct signal and a number of reflections from the ground and objects. The hemicap is used to represent cylindrical objects on the ground, such as fuel tanks, that might corrupt the direct signal.
"The objective is to measure the distance from the satellite to the receiver," Cohenour said. "Any signal that bounces off the ground or an object travels a longer path and corrupts the direct signal at the antenna.
"We can model the reflections from the ground and buildings, but we also need to model things like aircraft, fuel tanks and, of course, giant Tylenol capsules. We have shown that we can accurately model the ground and the buildings, now we are working on cylindrical objects."
The team's goal is to predict what the GPS signal is going to do regardless of the situation and regardless of what land features might cause obstructions in the signal.
"You can't just land an airplane with GPS because even though you might have sufficient accuracy," Cohenour said, "you can't guarantee that there's not some error in the signals. If the airplane's in the fog, and there's something wrong with the GPS satellite, it might run the airplane into the ground."
Cohenour has receivers on the ground at the test site that pick up the GPS signals. When conducting an experiment, he and Barker put an antenna on the ground and monitor the GPS satellite signals for three days. Then, they put the hemicap up, watch the satellite signals for another three days and measure the difference in error when the hemicap is present. They can then look at the measurement error, which are caused by the signals bouncing off the ground and other objects and off the hemicap, and tell what the changes are.
Specifically, they are using mathematical models to show the electromagnetic field caused by the objects, the antennas and the GPS receiver. The question they seek to answer: Does the electromagnetic field of the model reflect what you'd get in real life?
Barker became involved with the project in May of last year after he had been interning with the avionics department at the Ohio University Airport. He wanted to learn more about electrical engineering, so he figured it would be a good experience for him. "It's a chance to take some of the stuff we've learned in school, equations and stuff that is hard to actually visualize, and visualize what I'm learning," Barker said.
One of Barker's tasks is taking part of the mathematical computer model and adapting them to run specific satellite passes and geometries of the experiment. His work involved a lot of field experiments and computer programming, which includes some advanced math. Barker said he work can be challenging at times. "What I'm doing as an undergraduate is something a graduate student would be doing for a thesis," he said.
Barker said once he graduates, he would like to stay in the avionics field and to hopefully one day work for a company such as Boeing or Lockheed-Martin. This project, which is sponsored by the Federal Aviation Administration, is beneficial to Clint's goals because "just having an 'in' and knowing what those people expect from you give me that much more of a heads up," he said.
With all of the hands-on research Barker is doing with the hemicap project, hopefully to a potential employer, he'll be someone to bet all of their chips on.
Josh Blair is a graduate student in the School of Journalism