Originally Posted by _m:
“I think the best thing for EE to do would be to get rid of 2G completely and sell all of the M2M contracts to O2. Given O2's current lack of 4G in most places, they'll be needing 2G for a while longer whereas EE could just turn it off now (and put 4G on full power!) and most people wouldn't even notice.
As for devices without VoLTE support, they could either force them to upgrade, or just sell off their contracts to any of the other three - EE may lose some money from it but it'll be worth it in the end when they no longer have to spend money supporting legacy technology so they'll have more money to spend on building new sites to fill in any coverage holes thus having 4G coverage virtually everywhere.
With TETRA being decommissioned soon and the 700MHz auction right round the corner, EE could even move 3G across to those frequencies (and use 2100 for 4G!) and then keep all of the contracts for non-VoLTE devices as, even with cell breathing, they'll have 3G everywhere they have 4G, essentially using it as a backup and for calling on older/inferior devices.”
As much as I love this idea, it's just not going to happen.
I do think that getting rid of 2G would be good for data users, as they wouldn't end up being shoved down to 2G in areas that either 3G or 4G is weak (but still usable). Some phones do do this! 3G or 4G is a must for data use, and 3G is also recommended for phone calls, as the quality of 3G (and VoLTE) calls is better than that of 2G calling. So basically, whatever 2G can do, 3G can do better... with the exception of coverage, due to cell breathing. That's where 4G is much better, but it doesn't natively support voice. So non-VOLTE phones have a problem.
Switching off 2G would also result in more 1800MHz capacity for 4G, which would help in certain areas. As far as I am aware, 2G is using 25MHz of the 1800 band? Refarm 20MHz to 4G and leave 5MHz for M2M. Or just sell off M2M to Vodafone/O2.
Don't think that refarming the ex-TETRA spectrum to 3G would be very viable at all. Firstly, 3G@450MHz will definitely have a lot of raw coverage, and cell breathing wouldn't be as much of an issue. However, you're more likely to have cell breathing taking place - since 450MHz covers a vast area, then unless you have multiple masts covering the same area, 3G450 is going to get very swamped very quickly. If you've got miles upon miles of area covered by one 3G450 mast, then there will be an awful lot of devices on that mast. Not only will the mast be grossly swamped, but 450MHz isn't a brilliant band in terms of carrying large amounts of data. If you want high capacity transmissions for data, that's where bands such as 4G2600 come in handy. 450MHz won't have a lot of capacity at all - so even if it wasn't swamped, data would be slower than currently. But with swamped 3G450 masts, you'd be back to 2G speeds!
Oh, and of course, virtually no devices at all support 3G450, meaning that your devices become obsolete anyway.
The same goes for the 700MHz bandwidth - 700MHz will have more coverage than 2100MHz, but you've got more chances of the network being swamped. And again, there are very few devices that support 3G700 (though there are probably quite a few devices that support 4G700).
If you really want to keep 3G for the long term, your best bet would be to do a deal with Vodafone (or O2) for their 3G900 spectrum. That way, with the amount of masts EE has, using 3G900 at full power would resolve cell breathing issues, and you'd have coverage that either exceeded or matched 2G1800. Then all but 5MHz of 1800 could go to 4G, leaving the 5MHz 2G for M2M only. That, of course, would result in no 3G in a lot of areas on O2/Vodafone, but some areas have never had anything more than 2G from Vodafone/O2 anyway.
But even though my plan would be semi-good (more realistic than your plan, anyway), neither are going to happen. Sure enough, there may (or may not) be plans in the long term to decommission TETRA and use that for 4G, but that would have to be VoLTE only. As much as I dislike this (as it will make quite a few data devices obsolete), chances are 3G will go before 2G, and when VoLTE is the norm (on every handset), then 2G will finally disappear.
Short-medium term: 4G rollout completes, 800MHz rolled out network wide (along with VoLTE), possibly 2600MHz added to more masts (to deal with congestion)
Medium-long term: 3G2100 gets partially (or fully) refarmed to 4G2100, 3G switched off, power increased on 4G1800 to match 2G levels. This would mean that 2G has to deal with a lot more calls, so could only feasibly happen when the vast majority of devices are VoLTE capable.
Long term: Everything runs through Volte, M2M devices either get converted to 4G, switched to another network or switched off entirely. 2G gets switched off, everything goes through 4G (or even 5G if it is released).