Well, it seems all good clocks in low earth orbit will eventually succumb to the rather small effect the outer atmosphere plays on the retro rockets/ mass reactors, whatever.

One by one our GPS satellites are coming to ground. Mostly due to not being able to maintain elliptical orbits.

However , there are satellites up there. Most, although not designed for GPS, have some kind of CPU set at some kind of clock speed. Most are in known (read calculable) orbits. All have known clock speeds. All of them beam clock speed down as a matter of legacy principle (from the old days). Err. GPS?
-- 4whom, Jul 21 2009

Lights in the city of Satte http://en.wikipedia...wiki/Satte,_Saitama
[normzone, Jul 21 2009]

Clever. I wouldn't want to do the math though. Is clock speed that constant? Not being an electrical guy, I'm not up with that stuff.
-- RayfordSteele, Jul 21 2009


Sweet mouther of *** I love it, but I don't think I can do the math . Wouldn't you have problem with sattelites of close to the same age have the same clock speed?
-- dev45, Jul 21 2009


//All of them beam clock speed down as a matter of legacy principle (from the old days)//

could you explain further?
-- xaviergisz, Jul 21 2009


Most satellites will be coming down much sooner than the GPS birds will. (With the exception of the geosync sats, which are useless for GPS duties.)
-- lurch, Jul 21 2009


Could you provide a link to something that talks about the kind of information given by these satellites? The GPS works on the premise that one knows where every satellite was at the time of the (synchronized) signal - how would that work with just any satellite?
-- loonquawl, Jul 27 2009


GPS "satellites" (thanks [normzone]) are falling out the sky. They are coming out of the sky faster than other satellites because they have had to maintain orbits where at least four of the 24 are in line of sight from any particular point on the globe. Major strain, and against bits of atmosphere we did not account for until quite recently.

Other satellites have no such restrictions. However they do fly calculable orbits (well at least until their mass reactors run out, but that is far in the future, we could launch another three GPS programs in that interval).

These other satellites, besides having calculable and predictable orbits, also have CPUs. These CPUs have clock speeds, measured in Hz, rather unfortunately, but rather more fortunately accurate in their frame of reference.

In the case of imaging satellites the clock speeds are important, as they clock the "temporal resolutions" of the CCDs or CMOS image data collectors. In the early days, as a matter of principle, the clock speed of the CPU was sent as data in the header. This was because of the worry that time is a bit slower the further up you go. This data remains as a legacy, even though we can now calculate for it, in certain data headers. To be quite honest it was actually used as a parity mechanism in the early days (unfortunately).

And even where it doesn't exist in the header data it still is reminant in the imaging data we collect.

All I am saying is that we did need a few clocks accurate to one billionth to calculate position. These are falling out the sky, soon. Nothing is up there to replace them. Any CPU from 1 GHz upwards will give us some accuracy (and believe me they are up there) that we can use.

The maths is terrible, even worse, your current terrestrial GPS measure can't measure in hundreths anymore like it does now (no more fourth satellite operating at billionths), it has to be on a similar clock speed. *You* would have to carry a clock accurate to a billionth. As far as I can see the only thing stopping this idea is the creation of an X number of atomic clocks that fit inside a small mechanism for retail. That is probably harder than launching the next gen of GPS satellites. Ergo, fishbones welcome...
-- 4whom, Jul 27 2009


Current GPS satellites work by receiving their location *as transmitted from the ground*. They're literally just bouncing time and location that they receive from elsewhere.

Therefore, the code on the satellites up there would have to be rewritten to accommodate your idea. Tell me why globalstar wants to re-purpose their extremely complex satphone satellite to perform what amounts to the oldest, most rudimentary satellite function possible (other than orbit)

I have a very strong suspicion that the whole sudden "OMFG, SATELLITES ARE GOING TO FALL OUT OF THE SKY AND MY TOMTOM WON'T WORK AND I WON'T BE ABLE TO FIND MY WAY TO GRANDMA'S" is just someone's PR engine going into overdrive in an effort to get some extra cash/funding to spend.

Name one thing even half as popular/relied upon as GPS that disappeared from the market because it got old (things that turned out to kill people (asbestos, DDT) don't count)

If the system starts to falter, it will be repaired before anyone even notices it.
-- ericscottf, Jul 27 2009


//Current GPS satellites work by receiving their location *as transmitted from the ground*// That is just plain wrong. I presume the rest, written in capitals, is cut from the same cloth. It is just stupid. Do some reading on "How GPS works".

I don't really care if you can't find your matriach. I am a bit more concerned about A380's and Tunnel boring probes, that, "can like to rely" on GPS info, to do their jobs.

Your GPS device, in your hand, is simply an accurate quartz clock (nowhere near a thousandth, ito accuracy) and a database of known orbits and "callsigns". No information is given to satellites to get your position, as that would require about 186 ms of your time, skewed by the satellites time and then back down again, err, not good for accuracy..
-- 4whom, Jul 27 2009


//Do some reading on "How GPS works".// I'll throw that one right back at you.

//GPS "satellites" are falling out the sky// Sez who? (linky, please). GPS satellites fly 12-hour orbits, and are therefore *much* higher, and therefore *much* further away from atmospheric drag.

//they have had to maintain orbits where at least four of the 24 are in line of sight from any particular point on the globe.// This week, there are 30 Block 2/2A/2R/2R-M GPS birds active. (SVN54 was due to be out of service 7/25-7/26 for repositioning, should be back in by now.) With a single satellite, you can fix your position as being on a sphere centered on that satellite. Not very useful, as that will draw a circle on the earth's surface. With a second satellite, you get another sphere, which fixes your position somewhere on a circle marking the intersection of the two spheres, which in turn marks two dots on the earth surface. No altitude fix yet, but you might be able to guess which dot. With a third signal, you can determine which dot you're on. A fourth signal allows you to determine HDOP and VDOP - horizontal and vertical dilution of precision - and make a judgement of how accurate your fix is. Point being: four sats are only "necessary" for a precision 3-dimensional fix from a cold boot. If you are out flying around, you can still take 3-d fixes from 3 satellites, and you lose altitude reference if you drop to 2 sats.

//Other satellites have no such restrictions.// Where in the name of hydrazine did you come up with that one?

//In the early days, as a matter of principle, the clock speed of the CPU was sent as data in the header. This was because of the worry that time is a bit slower the further up you go.// Very wrong. A clock sent to high altitude won't be able to tell a difference in *its own* clock rate. Sending info to the ground about what it thinks its clock rate is would be useless - if it was a 1 MHz clock on the ground, it'll be a 1 MHz clock on orbit. (If someone on the ground were timing the clock, it would appear to be faster. That would have to be observed from ground, though; the on-board systems can't measure it. That's how relativity works.)

//Any CPU from 1 GHz upwards will give us some accuracy// You would have to know not just the rate, but the alignment - just *when* does the front edge of the signal occur *in relation to that received from a different satellite*. It's not enough to just chop a second into a billion parts - you have to know *which* of those billion occured at the same time as a billionth from another source. ("Synchronized", I'm trying to say. Those LEO satellites aren't, and don't have the capability to be, synchronized with each other.)
-- lurch, Jul 27 2009


Ok [lurch].

[ericst...] implied GPS satellites obtain information from ground based data with the statement //Current GPS satellites work by receiving their location *as transmitted from the ground*. // That is wrong. He should read up on "How GPS works".

//Sez who? // Sez everyone! (google it if you want a link) It is important to note that not all 30 satellites need to be inaccurate, only a fraction of them. I am not saying all 30 (or 24) need to drop out of the sky at once. Maybe I used the term "atmospheric drag" a bit loosely, there is certainly some action on the satellites from the earth other than gravity. These satellites have been fighting these forces actively, and are running out of fuel.

//With a single satellite, you can fix your position as being on a sphere centered on that satellite. Not very useful, as that will draw a circle on the earth's surface. With a second satellite, you get another sphere, which fixes your position somewhere on a circle marking the intersection of the two spheres, which in turn marks two dots on the earth surface.// In order to get a relatively good position on the sphere with an inaccurate clock (your terrestrial GPS unit) you need four accurate clocks (with known postion) to give you a position, the fewer clocks of known position the more accurate your terrestrial device needs to be. You could even do it with one, but then we are talking orders of magnitude greater than a billionth of a second, and some other physics.

//Where in the name of hydrazine did you come up with that one?// Perhaps I should have said *some* other satellites don't have these restrictions.

//A clock sent to high altitude won't be able to tell a difference in *its own* clock rate.// Yep that's pretty much what I am saying. As long as it sends what it thinks, we can calculate the change, because we know its orbit. As long as the clock is operating at Ghz, we have some degree of accuracy.

//You would have to know not just the rate, but the alignment // Unfortunately, this idea fails on that point. And for that I congratulate you.
-- 4whom, Jul 27 2009


I was under the impression that GPS satellites were made aware of their own orbital position from transmissions on the ground. Cursory search does not reveal this to be the case, however, I was unable to determine how the satellites do indeed ascertain their position, so I am not yet ready to abandon my theory that at some point, something on the ground says "hey, satellite F, you're located [[here]]. Also up for personal research is whether or not location data is constantly updated on the satellite from the ground, or computed on board.

I didn't ever mean to imply that I was under the impression that GPS receivers were anything more than passive. That is silly, though I do know several people who refuse to buy GPS units because "the satellites can see where you are and the government can track you"

Lastly, my original point: GPS isn't going anywhere. Name one other popular technological thing that disappeared permanently simply because part of it broke.

For someone who likes to deride others comments so quickly, I recommend you

1) Make more of an effort to explain your ideas in a grammatically correct way. 2) read people's responses properly so as to avoid interpreting their comments incorrectly.
-- ericscottf, Jul 29 2009



random, halfbakery