[Don_1 (OP)]

Geographically speaking, I am pretty knowledgeable about where I am, where I want to be and how to get there. But as I often have to go to some back street in a small town or village in the middle of nowhere, I have a sat-nav.

The other day, I had occasion to use my sat-nav to find an address. I was stationary (eating a nice bacon sarny for breakfast). The sat-nav showed my speed ranging from 0mph to 2.6mph over a period of about half an hour. At one point it went from 0mph to a mind-blowing 3.4mph!!! But I was stationary! It's not the first time this has happened.

Since my sat-nav pinpoints my position by triangulation with satelites which are in a geostationary orbit, I wondered just how geostationary these satelites are.

_{Mod edit - Formatted the subject as a question - please do this to help keep the forum tidy and easy to navigate - thanks!}

[DoctorBeaver]

I doubt it was due to the satellites. More likely is atmospheric disturbances of some kind interfering with the signal.

[syhprum]

The satellites are not geostationary but orbit about twice a day

here is a very good description of the system

http://en.wikipedia.org/wiki/Global_Positioning_System

[DoctorBeaver]

Well, that shut me up  []

[LeeE]

The accuracy of GPS receivers also depends on the location in the sky, relative to you, of the GPS satellites.  GPS receivers seem to be, from my own personal experience with a hand-held trekking unit i.e. not a driving unit that always locates you on roads, most accurate when the sats you get fixes from include a couple at different points on the horizon, in addition to the ones overhead, which makes sense to me as you then get the best triangulation.  However, because the stats that are low on the horizon are either going to dip below it, or rise higher, you'll either lose their signals, or their triangulation will become less optimal.  My hand-held unit actually shows the sat positions and you can watch the estimated accuracy of the fix varying as you track the rise, transit and fall of the sats.

[lyner]

If you were in your car then there may have been vehicles passing you. They disturb the timing of the signals because they cause multipath propagation. Park out in the open, away from traffic and try again.
The GPS on my boat shows me wandering about whilst at anchor (which we do) but the overall accuracy is not far from what it claims to be.

[LeeE]

You should normally have a couple of good horizon fixes on a boat.  On land, unless I'm very high up, the horizon is usually quite a bit higher than me so the estimated accuracy tends to be lower and at times can be a couple of hundred feet out.  It can even be difficult to get a four-channel lock in built up areas where you're surrounded by high buildings as the GPS can only see a relatively narrow cone upwards.  Heh - the elevation then tends to be quite accurate though, even if the lat/lon are poor.  In practice, it's rare for me to get a twelve channel lock, which is the limit for my hand-held.

Sophie - does your boat GPS use differential mode?

[Don_1 (OP)]

Well, that shut me up  []

Geeze....... I'm not surprised Doc! It's all a whole lot more complex than I had imagined. There I was thinking these were geostationary satelites and a spot of straightforward triangulation.

Just goes to show you should never assume anything. Thanks for that Syhprum.

[graham.d]

You need multiple satellites and a reasonably clear line of sight to them for correct operation. Usually the receiver will tell you how many satellites are being received. There is no angular measurement as the signals are simply transmitted over a wide angle and there is no directionality in the receiver either. The position comes from the relative timing that signals are being received from each satellite and the satellites knowing where they are. There is a rolling pseudo-random code transmitted from each satellite which is sent with very, very precise relative timing (atomic clock precision). A good description is on wikipedia:

http://en.wikipedia.org/wiki/Global_Positioning_System

Lee, I don't think differential gps is used on boats in the uk (at least not on pleasure craft). It relies on a positionally fixed point nearby to act as a reference which is not usually practical at sea although can be useful in sailing into harbours or other hazardous areas. I think the USA does have a system where error corrections are broadcast on VHF frequencies. It is not really necessary for small craft as the accuracy is usually sufficiently good anyway. You are right about the need to have a good spread of satellite positions. The problem with satellites near the horizon is the atmosheric effects and their variability. Differential GPS can help here.

A great system for boats is AIS. This uses the GPS system then broadcasts your position and vector (and loads of other useful data) on a VHF band so it can be reconstructed on receiving vessels' chart plotters. It is compulsary on vessels over a certain tonnage but receive only systems are very cheap to buy and are a worthwhile investment for smaller vessels.

Just as a further comment on differential gps, it can be very accurate and is used more and more in surveying and in construction. Accuracy can be as good as a few millimeters on the more expensive systems.

[syhprum]

I think the differential system was mostly used when the unencrypted signal was deliberately degraded to prevent 'rogue nations' using it.
They found this caused as much trouble to the goodies as the baddies and now it has been abandoned

[LeeE]

I thought differential GPS often covered coastal waters, obviously not the open sea or ocean, but then I'm not a sailor []

[graham.d]

Lee, I think you are right, having now actually checked up on my previous assertion :-)

http://en.wikipedia.org/wiki/Differential_GPS

I have to say, although I am a sailor, I did not know this was ever implemented in the UK. I guess it just is not relevent to amateurs like me and normal GPS is so good now anyway. Maybe it is as Syphrum says. It would not seem very necessary now.

[anthonyy]

I use telenav ( newbielink:http://www.telenav.com [nonactive] ) on my blackberry, and was also wondering the same question.  How come some GPS providers such as TeleAtlas claim to have better reception? 

[lyner]

It sounds a bit scary to know that the system is getting creaky. I imagine that the main problem will be for fast travelers because there will be times when a passing satellite is not working. If mine works OK every ten minutes, then it's fine for travel at 6kts max.

As for differential operation at sea, the antenna is seldom mounted higher than a couple of metres above the sea so the reception horizon will be extremely limited and the signal would be subject to chronic multipath interference.