What happens when wireless spectrum runs out?

[gotapresent]

This question might be better answered by someone in the mobile industry, but I guess I'll ask it here since it does involve physics.

I understand that wireless spectrum for mobile data is finite, and will run out at some point. Out of curiosity, what are the implications of that exactly?

Does that mean that there's a hard physical limit on how fast future mobile networks can be, or how many subscribers they can support? Does that mean that there won't be '6G' or '7G' etc networks in the coming decades?

 

[skeptic2]

The limited spectrum is an example of the Tragedy of the Commons. It can be dealt with in a number of ways. One is that it can be auctioned. As the spectrum fills up the price will skyrocket which will mean poorly funded but very important functions may not get spectrum. Think about municipalities trying to get spectrum for police and fire on a limited budget.

Another method is to grant frequencies on a first come first serve basis and when there are no more frequencies, nobody else gets any.

There's not really a hard physical limit on the number of subscribers. As the frequencies disappear, power could be reduced in order to allow more co-channel reuse. As more bandwidth is used, the number of base stations may need to be increased also to increase channel reuse. For applications where immediate reception is not needed, packet transmission can be used.

 

[DaveC426913]

That's why we go digital.

Analog communications use different frequencies to separate signals. Number of users is limited by how many times you can split up the bandwidth.

Digital communications use the same frequencies but encode them uniquely. Number of users is limited only by number of unique cyphers you can produce (there is no upper limit - of you need more cyphers, you just use a bigger number - rate-limiting factor is merely transmission rate, which is muuuuuch more forgiving than bandwidth). Digital devices can hear everyone else's signals but ignores all but the one they can decode.

 

[f95toli]

Number of users is limited only by number of unique cyphers you can produce (there is no upper limit - of you need more cyphers, you just use a bigger number

This is not quite true. Even with a digital stream you still need to keep track of where the information is suppose to go and this gets harder as the available BW goes down.
3G and 4G mobile communication systems need something called multi-used detectors which are effectively DSP htat "decode" the stream (to avoid interference). This is unfortunately a hard problem (in the computational sense) and can not be done efficiently using parallel processors; i.e. a SINGLE processor per tower is effectively doing all the number crunching.
Hence, one reason why the your data transfer rate goes down when there are many users using the same cell tower is that the MUD can't keep up. The only way to increase capacity is to build more towers or use a faster MUD.
Fast MUDs is one application of very fast electronics (I believe some of them are even implmented using III-V technology, Si is not fast enough), and the DSPs are already VERY fast. The question is of course how much faster they can get. The answer to that question is probably "not much".

 

[wuliheron]

Its less of a tragedy of the commons and more like the wild west waiting for the invention of barbed wire to tame the countryside. The airwaves will always be there and there's plenty to go around, but for now we can expect range wars. That's how new frontiers are developed in capitalistic societies and its assumed civilization will eventually follow. I just wouldn't hold my breath waiting for it anytime soon in the telecommunications industry. It will likely be a few more decades before the technology finally catches up with our ambitions.

 

[skeptic2]

Its less of a tragedy of the commons and more like the wild west waiting for the invention of barbed wire to tame the countryside. The airwaves will always be there and there's plenty to go around, but for now we can expect range wars. That's how new frontiers are developed in capitalistic societies and its assumed civilization will eventually follow. I just wouldn't hold my breath waiting for it anytime soon in the telecommunications industry. It will likely be a few more decades before the technology finally catches up with our ambitions.

My job is managing spectrum and I couldn't disagree more strongly.

 

[DaveC426913]

My job is managing spectrum and I couldn't disagree more strongly.

Good. I thought maybe I was wrong. We are running out of bandwidth. There are ways around it, but the squeeze is on.

 

[gotapresent]

Thanks for the replies, very informative.

If I understand right, that means that there's still some spectrum that's unused but poorly distributed, and that there's still room for further optimizations to mobile network infrastructure. How soon do you think we'll reach a point where neither of those are true anymore?

 

[f95toli]

I think it depends.
Firstly, I don't think it is true that there is "free" BW anywhere in the part of the spectrum that is easy to use; all of it has been allocated for a long time in most countries and the only way to introduce new services is to close down old one. The fact that 4G will only be introduced here in the UK AFTER the analogue TV network has been closed down in as good example.

However, one can of course go to higher frequencies but the problem is cost. Technologies like WiMAX etc. can use higher frequencies that haven't really been used much for consumer electronics, but the drawback is that microwave componentes (amplifiers etc.) strt to because very expensive and have to be build using costly fabrication technologies.

 

[wuliheron]

My job is managing spectrum and I couldn't disagree more strongly.

We could install a dozen different short range transceivers in every lamp post. We aren't running out of bandwidth, we're running out of cheap bandwidth and still auctioning off the rights to the cattle barons. If you manage the spectrum then its your job to turn a blind eye to the politics as the commons are increasingly encroached upon by the oligopolies.

 

[DaveC426913]

However, one can of course go to higher frequencies but the problem is cost. Technologies like WiMAX etc. can use higher frequencies that haven't really been used much for consumer electronics, but the drawback is that microwave componentes (amplifiers etc.) strt to because very expensive and have to be build using costly fabrication technologies.

You're still talking about microwaves though. They are still limited. There is an upper limit. You will start getting into IR and visible light. Problem is, these frequencies don't have the same penetrating power so you can't use them the same.

 

[wuliheron]

You're still talking about microwaves though. They are still limited. There is an upper limit. You will start getting into IR and visible light. Problem is, these frequencies don't have the same penetrating power so you can't use them the same.

Bandwidth refers to broadcast range as well as frequency. I could build a CB radio that would reach the opposite coast, but the thing is illegal for a reason. We have plenty of microwave bandwidth to go around, especially if you limit broadcast ranges further, and the only real issue is that the big telecoms are approaching the limits of what can cheaply and easily be exploited without a lot of regulation.

I'd compare it to Netflix or bittorrent taking up internet bandwidth. There are ways to expand the internet bandwidth, but it costs money. There are ways to watch movies without streaming them over the internet, but it costs money. The issue isn't a tragedy of the commons, but a simple problem with large companies competing for cheap resources in a largely unregulated market.

 

[f95toli]

You're still talking about microwaves though. They are still limited. There is an upper limit. You will start getting into IR and visible light. Problem is, these frequencies don't have the same penetrating power so you can't use them the same.

Sure I am talking about microwaves. I can't see how you could ever use frequencies much higher than a few tens of GHz for any meaningful wireless communication standard, as soon as you reach say a 100 GHz or so you enter the realm of quasi optics where the radiation essentially behaves as "beams", a bit similar to visible light (I've done some work in the 100-
1000 GHz range, and mostly you end up using quasi-optical lenses to direct the radiation).

Hence, if you wanted to use these frequencies for wireless communication you would essentially need a direct path between emitter and receiver. Another problem is that many frequencies in this range are attenuated quite heavily by the atmosphere (water, and rotational absorption spectra of various molecules).
Note also that we simply do not have any efficient technology that can operate in this regime, everything is very bulky and extremely expensive ("traditional" equipment only operates up to about 110GHz, but everything about 65 GHz or so costs a LOT of money).

 

[HowlerMonkey]

Hopefully, this will make downloading family guy episodes on your smartphone super expensive and we all can get back to work.

 

[Dickfore]

Cell phone base stations recycle the same frequencies beyond nearest neighbor cells, don't they? I thought that is the advantage of GSM. But, perhaps you were referring to something else.