Lab Report: GPS Error

Michael ShortPrincetonUniversity

Where’s Waldo?March 8, 2004

Lab Report: GPS Error

PURPOSE: The purpose of this lab was to determine the accuracy over time of a GPS receiver standing still. To measure this, we were to record the data received by the GPS handheld thing over the course of an hour.

DOING IT: Well, when Sunday came around and it was time to do the lab, things seemed perfect. It was a beautiful day outside, and people were sitting in the grass everywhere reading, studying, and doing work. Nick and I though we could go out, computers in hand, and not only accomplish our GPS tasks for the week, but really accomplish some genuine roommate bonding at the same time. Alas, things were not to go as expected. Just before heading outside, I realized that I had lent my computer connecting cable to someone in the class with the basic Etrex model, and so Nick and I only had one cord between us. We would have to each record data at a different time. However, being the nice guy that I am, I didn’t want to make Nick go outside and sit there all by himself, so I went out with him anyway, thinking that I would just record data as soon as he finished. It was only as his data collection was winding down, and the sun was setting, and dinner time was approaching, that I realized Nick would not return my courtesy. He fully intended to leave me all alone as soon as he finished. In protest, I refused to do my own lab. In fact, since I had figured out how to get the computer to record the stuff in the first place, and had actually physically set his computer to do so, I had really done his lab. If anyone should have had to sit outside alone, it should’ve been him.

CLASS MONDAY—ACTUALLY DOING IT: Unfortunately, the data was indeed on Nick’s computer, and since ownership is 90% of the law, I had no lab when I walked into class Monday. Fortunately, several other classmates had failed to retrieve data as well, so Prof Groth let us go outside for the first hour of class and get some solid info from the orbiting gods of position. Thankfully, it was another nice day, and the four or five of us hoarded GPS signals from above to the mellow sounds of Garth Brooks’ guitar.

USING THE DATA—THE FINE ARTS OF SPREADSHEETING: Oh boy is spreadsheeting fun. This time we delved into the marvels of standard deviations to see how consistently the GPS calculated our position over the course of the hour. As always, all the various interesting formulas I used are visible on the spreadsheet that is also attached. My data seemed pretty accurate overall; the standard deviation for latitude, longitude, and altitude were all quite low. That’s likely because I often had a fix on seven satellites, instead of the minimum four. So, yeah, take a look at the data. It’s pretty self-explanatory.

GRAPHS + WEIRD STUFF: What could you do with so much wonderful data, if not graph it? SO I graphed it. I wanted to see how the error levels changed over time, so I graphed the standard deviations of all three measurements of position over…well, time. Check out the graphs. Bounces back and forth quite a bit, eh? So it goes. Of course, the graphs look strange. My graphs always look strange, for some reason, and that reason is that I usually screwed something up. Thankfully, problems are soooooo obvious on graphs that my errors become inescapably apparent. This time, it appears that there were tremendous gaps of time over which no data was being recorded. I had actually been wondering if that might happen. You see, I noticed several times while recording data outside that my computer put the screensaver on, and once even went into “lockdown” mode. When I made it come back on though, the GPS program was still running, so I thought perhaps it had continued to record data. Well, I guess not.

CONCLUSION: The GPS is pretty accurate. It had me basically to within a 2 meter radius in terms of my horizontal position, and within about 4 meters with respect to my altitude. That’s what the standard deviation was, anyway. I approve.