If we accepted some loss of accuracy and coverage, would it be possible to have GPS-positioning without making use of satellites? If so, how might such a system work?
Sure. I did a search on something I remembered reading about, but here is an article on DARPA making a non-space based GPS system.
I seem to recall that one of the big companies (Google maybe) was looking into using tethered balloons or maybe balloon trains to do something similar (as well as provide wide scale wireless internet and cell service), so it’s possible. How feasible it is would be another question.
If it’s not global, and there are no satellites, you basically have LORAN. Works using multiple terrestrial radio beacons.
Or you could always just pull out the old sextant. (No loss of coverage, unless it’s cloudy!)
LORAN was essentially a land-based GPS-like system. Dozens of stations transmitted signals which were coordinated with each other to high precision. For example, if you received a signal from station X one microsecond later than the signal you received from station Y, that meant you were about 984 feet closer to station X than you were to station Y. This placed you somewhere along hyperbola. If you could also receive a signal from station Z, that would give you two more hyperbolas. Where all three intersected would be your location. The accuracy of the system was about +/- 60 feet. LORAN was discontinued right about the time GPS was declassified.
The advantage of having GPS in space is that you need only a handful of satellites to service the entire planet. You could have every cell tower broadcast its precise location and that would work for GPS–as long as you are within range of multiple towers (just one or two won’t be enough.) So it would work only in highly serviced areas that currently have electricity.
Or google “pseudolites.”
How did the accuracy of LORAN compare to GPS in its early days? How about GPS now?
What was the range of the stations?
Could a LORAN-like system be supplemented with airborne relays? If airborne relays their location based on land-based stations and you can get the signal from 3 airborne relays, then you know where you are (with some loss of accuracy).
I’m surprised a system like that isn’t in place in case the US gets into a war with a nation which has soft-kill or hard-kill ASAT weapons.
I don’t understand how the GPS system works, but I thought part of the system operation relied on the doppler shift of the radio signal. I thought that was one of the reasons it can be so accurate.
If not, just let me know gently… 
Your phone actually does determine its location, to some extent, this way, based on what cell tower(s) are visible and with what strength. If it’s Android, it also determines it to some extent based on what WiFi networks are visible to it, and where those are known to be. It’s not as exact as GPS at its best, but it’s quicker and more reliable.
Not quite. A doppler shift would register the speed something is coming at you or going away from you. GPS satellites have very accurate clocks, which transmit the time very precisely. Knowing the position of the satellites, and by measuring the small differences in the time on each of the signals, one can arrive at a very precise location on earth.
It’s probably in the link DPRK gave, but basically and from memory, it was pretty expensive, difficult to maintain and also not nearly as accurate. It wasn’t nearly as scalable either.
I think if you were to do a new, modern non-satellite based system you’d almost certainly do an airborne one (IIRC, LORAN just used radio navigation, don’t recall what the acronym meant, but you could do a radio based system on anything). I recall there were several vendors looking into balloon based systems (China was also looking into something like this, again IIRC) or long duration drone systems, and that would be more scalable than ground based systems…also they would have more utility. Hell, they might be better from a deployment and maintenance perspective if you had the bucks to bankroll the initial capital expenditure and leverage that with usage. The system we have is useful because, well, we have it and someone (Uncle Sam) footed the bill for it. Doesn’t have to be that way in the future, though satellites have obvious advantages.
Well, several of the LORAN variants did actually continue on through the 70’s or 80’s I think, so we did actually do that. It just wasn’t commercially available. Then again, neither was GPS initially, so there is that.
GPS in its early days was just as accurate as it is now, as long as you had a military GPS receiver.
Many people don’t realize it today, but the original GPS system was built for military uses, the US military in particular. They gave us access to a much less accurate civilian signal. Basically, IIRC, they skewed the GPS signal to be off by a little, and the information to correct the skew was contained in a separate, military-only signal from each satellite.
The skew amounted to a position error of around 100 meters, which is handy but not accurate enough for many uses such as precision navigation or Geocaching. Seeing its applicability to these, President Clinton ordered the military to set the skew to zero, until such time (like war) that the military really needed everyone else to be less accurate.
ETA anecdote: I went to an IEEE meeting many years ago, where the topic was GPS. The speaker had a GPS receiver, about the size of a large shoebox, and announced during the presentation that just after 10:00 PM there would be enough satellites overhead for it to determine a position (it needs four, and the full constellation hadn’t been launched yet). So at the end of the talk we all went out into the parking lot and watched it do its thing. Exciting stuff.
VLF is being considered in a GPS denied environment.
https://rntfnd.org/2016/06/17/darpa-eyes-very-low-frequency-signals-for-pnt-systems-as-gps-backup-govt-executive-magazine/
A couple of points to that. First, the GPS technology has actually improved since the early days. The civilian intentional degradation was there, but even before that was turned off, the makers of civilian devices were looking for ways around it. You can get some improvement by using the phase of the radio waves themselves, in addition to the signals carried on the waves, and you can also get improvements by mounting GPS devices in fixed locations with known coordinates (like cell towers, nowadays), measuring how far off they are, and using that as a correction for other nearby devices.
Second, I’ve heard that the first time that the intentional degradation was turned off, it was because we were at war. The military had a hard time getting enough military-grade receivers quickly enough. They could, though, get enough commercial off-the-shelf ones, and it was judged that the enemy (with less access to American shopping venues) couldn’t get a significant number, and so it was tactically better for us to turn off the degradation, have all of our guys get precise data, and let a few of their guys get precise data, than to limit the precise data to just a few of our guys.
Could ionosphere-bouncing radio also work for non-satellite GPS? Over-the-horizon radar - Wikipedia
How about ground wave radio? Over-the-horizon radar - Wikipedia
Skywave and ground wave radar have very wide beams but in a GPS system, all that matters is triangulation based on delay, right?
Yes, the Omega VLF system and the Russian version work this way. Not as precise as GPS or LORAN, but you get coverage 10000 km from the tower.
ETA Wiki claims that precision is/was 2.5-7 km, but I wonder if that represents ideal or typical conditions.
Just for completeness, GPS has 2 resolutions, the C/A (coarse/acquisition) and P (precision). When SA (selective availability) was in use the C/A signal was dithered in time, which had the effect of de
grading the spatial resolution to worse than 100m. Even with C/A enabled differential GPS systems could correct for the dither, and achieve very high resolution. As noted above, Clinton acted to have SA remain off, and we are told that the latest block of satellites don’ even have the capability.
The P channel is a separate issue. The biggest problem with GPS is the effect of ionospheric delay on the signal, and since the ionosphere isn’t a constant (it varies over the course of a day and is significantly affected by solar radiation and solar storms). So you need a dynamic way of determining the delay. The ionosphere is dispersive, that is, the propagation delay is dependant upon frequency. So transmissions of two separate frequencies can allow you to work out what the absolute delay is. That is what the P channel does. The P channel is encrypted, and only military users can decode it, and this has always been so.
However the European Galileo system will provide free access to a signal that has even better resolution than the GPS P channel. Something the US were not exactly pleased about.
Difficulties with radio propagation delays are going to be the bane of any navigation system. Satellite systems have the advantage that they can place any transmitters in the sky with a clear path to your receiver. Terrestrial systems have no such luck. Multi-path propagation will degrade the resolution significantly, and can only be ameliorated so much by smart algorithms. For glabal coverage you have a host of dreadful problems. If you need global coverage you are going to need propagation over the horizon. That needs low frequencies and is subject to significant error. IE, Loran.
For cities and generally populated areas we already use cell towers, with most smart phones relying on cellular assist to enable them to get a fast fix on position. The problems with poor visibility of sky and cheap GPS units in phones has also lead to the use of mapping of WiFi MAC addresses as a way of adding another layer of precision. But you can’t navigate on the ocean or in the sky with these.
I was a LORAN Technician and Instructor of LORAN C Timing And Control while in the USCG. LORAN C advertised 50’ repeatability, that is if a fisherman recorded a LOR fix when he placed a trap, if he followed LOR fix back 3 days later, he’d be within 50’ of the trap.
You could get 500 N.M. for a station. Depends on which transmitter you had. You need at least 3 transmitting stations. The ideal chain had 4 or 5 stations; a Master and 3 or 4 Secondaries.
Airborne relays would NOT work. The transmitting stations NEED a fixed location.
You’re right that a generic moving beacon wouldn’t work. But it could be worked around.
Ultimately, like is surveying, an end-user’s position fix is as accurate as knowledge of the location of the base station(s), plus their knowledge of the geometry from the base to themselves. IOW "if I measure <this> distance (+/- ??) in <that> direction (+/- ??) from <known Point A> I must be at <point B> (+/- ??*??).
By making the stations fixed, like ground based radio beacons you reduce that source of error to essentially zero. So all that’s left is the problem of measuring your relationship to that fixed point accurately using time, signal delays, signal phase, whatever.
GPS satellites are most definitely not fixed; they’re moving at orbital speeds all the time. What they have going for them is very high predictability. Like anything else in orbit, if you know where it is at any given moment you can predict forwards or back for several orbits with real high precision. The GPS control center measures the location and orbit of each satellite very precisely and sends that data (You, Mr. Satellite, are <here> <now> and your orbit evolves like <this>".) up to the satellite which then re-broadcasts that to the world over and over as “I was <there> <then> and my orbit evolves like <this>; you do the math to see where I’ll be at the moment when it matters to you. Then figure your location relative to mine at that time.”
If an airborne beacon had a way to fix its position accurately, it could do effectively what GPS does: announce “I’m <here> <now>. Measure your own location off that.”
So there might be ways to help the airborne beacon determine it’s location that works precisely & reliably enough, but is so costly we can’t afford to equip every cellphone on earth with that system directly. So instead we put the expensive system on the aircraft or balloons or whatever, and have them rebroadcast their cheap signal to the cheap universal receivers in everyone’s phones, car, surveying stations, or whatever.
Back when GPS was just starting and still military, we used it for surveying when they financially had 4 satellites up. We would be out at a lot of different times when the satellites were all far enough above the horizon to be useful and we had only so much time to get readings and we would have to sit on a point with the receiver for up to an hour to get good enough data. This was all tied into one receiver set on a HARN marker which had a very high precision set of coordinates.
We were one of the first mapping companies to use airborne GPS camera positions and survey control & GPS firing of mapping camera positions. As satellites came on line and when the ‘jitter’ was removed it industry of mapping exploded with innovation.
Was flying out of Tulsa at time and we had LORN in the plane. At that location the LORAN coverage was very good and I set up many points to use as a poor man’s precision approach to some runways and airfields I would occasionally need to get into under less than ideal VFR conditions. Many would put me right within 10 feet of the center line and 10 feet from the threshold every time. So I had an EMERGENCY little black book for these. Only used them a few times but if you were always checking, as I did, as I would fly around the country, I found many places where LORAN was very accurate and where it was not. One of the reasons I am now an OLD pilot.
In the every day world today people expect a ‘zero error tolerance world’ as far as position is concerned and have quit thinking for themselves as can be noted by all the reports of people blindly following their GPS to a “LOST” position. Bawahahaha