Date: Wed, 03 May 2000 21:21:26 -0600 From: Warner Losh <imp@village.org> To: Matthew Dillon <dillon@apollo.backplane.com> Cc: hackers@FreeBSD.ORG Subject: Re: GPS heads up Message-ID: <200005040321.VAA36306@harmony.village.org> In-Reply-To: Your message of "Wed, 03 May 2000 20:03:48 PDT." <200005040303.UAA66590@apollo.backplane.com> References: <200005040303.UAA66590@apollo.backplane.com> <200005032313.QAA65552@apollo.backplane.com> <200005031744.KAA63550@apollo.backplane.com> <20000503200006.A35116@cichlids.cichlids.com> <200005031957.NAA01354@nomad.yogotech.com> <20000503130759.A15403@orion.ac.hmc.edu> <20000503221528.A37472@cichlids.cichlids.com> <20000503151513.D337@beastie.localdomain> <200005040230.UAA36034@harmony.village.org>
next in thread | previous in thread | raw e-mail | index | archive | help
In message <200005040303.UAA66590@apollo.backplane.com> Matthew Dillon writes: : SA can *not* be averaged out. People... SA is NOT NOISE. I will : repeat that. SA is NOT NOISE. SA is an intentional error : introduced by the satellites that looks like a drunk walking around : the town. It takes a lot of readings over a long period of time : (read: at least a couple of days) for any sort of averaging to : yield a worthwhile result. What I was trying to say was that SA is caused by the satellites reporting times that have a small offset added to or subtracted from them. Knowing where you are requires that you know what time it is to a very precise degree. Once you know what time it is, you can know where you are. That's why SA injects a pseudo random noise factor into the timing information that the satellites report. If you have an atomic clock and a GPS clock, you can measure the offset between the two fairly easily and graph the results. That is what I mean when I say you can compensate for the SA if you have a good atomic clock. If you watch a system that is measuring this, you'll see that one second you are 65ns slow. The next you are 64 ns slow, the next you are 68ns slow the next you are 65ns slow. Come back an hour later and you'll find you are 135ns fast, then 130ns fast, then 140ns fast. After lunch it might be 12ns slow, then 10 ns slow then 8ns fast. Just before leavinvg for the day, you might be back up to 165ns fast. All of this is relative to an independent, high accuracy time source like an atomic clock. That's why you can't take the average of 20 readings in a row. You'll just be factoring out a very small part of the SA sequence. There's some jitter in the SA, but its biggest impact is the long period error that it introduced. Put another way, it takes days for the drunk to complete his random walk. :-) Averaging over a few minutes will only tell you where the drunk is at the moment and not his true center. Which is what I think Matt is trying to say. Warner To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200005040321.VAA36306>