That was an explicit goal of OpenBTS when I was following it years ago. The GSM vendors want something like $250k per base station, which is unreasonable for third-world or very-rural applications. OpenBTS thinks they can get the radio hardware costs down to a tiny fraction of that to serve low volumes, backed by a Linux box running this software.
Interestingly, their first large-scale field tests were at Burning Man (http://openbts.sourceforge.net/FieldTest/), where there were thousands of active GSM handsets but no cell coverage. They were able to provide limited SMS support between local participants and in later tests allowed some outgoing VoIP calls.
The answer is to work with legal spectrum owners (aka cellular carriers), which is what we (https://fairwaves.co) and other in the space are doing.
Someone mentioned that building a full networks (or even a full base station) is much harder than just apt'ing some software and snapping some cheap SDR. So we spent couple years developing an SDR which is more expensive than your typical hobbyist one (http://umtrx.org) - btw even hardware of the simple version of our SDR is open-source - and then couple years building a base station which complies with all regulations and has open-source in its heart (based on Osmocom obviously). And now we've added a bit of (unfortunately proprietary) glue and we can run fully fledged cellular networks for carriers - mostly in developing countries, as was mentioned before, because demand for simple voice communications is through the roof there. At the same time you can buy one of our stations for your lab and play with it as much as you want (get your test license first though).
So yes, you can't stop progress and technology gets cheaper and more accessible. Which is very exciting.
The price for BS alone is down below $10k (afaiu, LTE oder Multi-system BS are at $20k). But to run this as a system you need more boxes than the BS only.
Sorrrrrrta. GSM the spec requires that 45 mhz of tx/rx spacing by convention, but GSM the technology as we colloquially know it, e.g. if you're code-shifting 2G (i.e. via CDMA) you could get away easily in the 900-928 range @ full duplex. No technical limitation at all there. This was briefly discussed in an older thread by myself and a few other chaps here[1] but I guess tons of other people have had the idea before too haha.
The US spec GSM phones operate at (a nominal) 850 mhz, but like you said there are plenty of eurospec (and asian spec.. and US issued quad-band..) which can operate with off-the-shelf with all the (again, nominal) 900 mhz[2] band tech that's being phased out. It will pick up, negotiate the handshake and route just fine. The issue is really just like any other cellular network: getting tens of thousands of antennas up, and having people buy new phones that are quad-band and/or carrying around an eBay phone for what amounts to a novelty.
I'm not on the cutting edge of milliwave or RF anything (all voodoo to me) but as I see Agilent come out with more and more sensitive test gear, I can only presume that the wireless industry is getting way better at tuning YIG's, VXCO's, eliminating side-band tone bleed, and what not. Much like the analog (i.e. 1g, remember those Motorola flip-phones? I do1!) -> 2g shift increased capacity by 3x, I'd have to imagine new encoding techniques and what not are being developed by the PhD's and postdocs. Who else is going to buy a $300k network analyzer otherwise, haha
[2] If you talk to old-time HAMs they'll joke about how the FCC really regrets throwing away a few hundred billion dollars worth of band in that spectra alone. You can pack a lot of content into each Hz down down there.
Could you clarify a couple of acronyms - YIGs, VXCOs?
Also isn't the real viability issue with this the MSC? Once you've registered with a base station you need to backhaul it to a switch that's capable of routing the call to another hand set. Maybe there are some good affordable options now for a soft swtich that also provides HLR/VLR capabilities? A real class 5 switch like those made by Metswtich are hundreds of thousands of dollars I believe.
Good point. Doing that would require at least a firmware change on the handset as I suspect the baseband code is built around the ARFCN (Arbitrary Radio Frequency Channel Number) which enforses that separation. It probably wouldn't be that hard to change the ARFCN formula on a handset, though the easy way would make it not compatible with regular networks.
The cost is down into hobbyist range. Not the cheapest hobby, but not a big deal for a lot of people. Look up USRP, HackRF, and BladeRf. I think BladeRf is the cheapest at $420.00 USD.
Putting up a single, limited range BTS is not the same as full scale, functional GSM network. It's similar as creating your home network with OpenWRT flashed router and calling it Internet.
Interestingly, their first large-scale field tests were at Burning Man (http://openbts.sourceforge.net/FieldTest/), where there were thousands of active GSM handsets but no cell coverage. They were able to provide limited SMS support between local participants and in later tests allowed some outgoing VoIP calls.