Obtaining US 5G supremacy is easy as Pai, says FCC commish Brendan Carr. It's all in the spectrum

Establishing US supremacy in 5G will require the release of more spectrum, FCC commissioner Brendan Carr claimed in a speech to the American Enterprise Institute this week. In his talk [PDF], delivered remotely due to COVID-19 restrictions, Carr bemoaned previous failings in infrastructure deployments and spectrum availability …

[Fazal Majid]

Pentagon

A big part of the problem is the huge swathes of spectrum squatted on by the military, which is why US carriers went all-in on mmWave dead-end frequencies that can literally not punch their way out of a paper bag.

[DS999]

Lack of spectrum is not a problem!

It isn't like the number of smartphones in use in the US is going to rise much more, the only thing that can rise is their data consumption. Higher data limits would push people off public wifi and onto cellular for more stuff, but there's a limit to how much data people can consume per unit of time - most popular social media apps and websites offer HD video. Hard to go much higher in data consumption that that, since phones don't have 4K displays and even if they did it would be almost impossible to tell the difference between native 4K video and upscaled 1080p on a 6" display.

Simply refarming retired 2G/3G spectrum to mixed 5G/LTE usage (the standard allows for the same swathes to be used for both) would satisfy most of that demand. The addition of stuff like 600 MHz and the millimeter wave stuff for density in airports, stadiums, etc. can satisfy the rest. There's no need for another gigahertz wide block at 6 GHz, military radar frequencies at 3.5 GHz and so on.

There is already a ton of licensed spectrum that has never been used - Dish Network is the biggest spectrum owner out there aside from the big three cellcos, and they have yet to put any of it to use!

[Henry Wertz 1]

"Simply refarming retired 2G/3G spectrum to mixed 5G/LTE usage (the standard allows for the same swathes to be used for both) would satisfy most of that demand."

It doesn't, Verizon is down to 2.5mhz (1 channel of 2G CDMA 1x, 1 channel of 3G EVDO), and running dynamic spectrum sharing (so the 4G and 5G spectrum are already shared on several 4G bands; not band 13 for regulatory reasons). AT&T sunsetted 2G quite a while ago (and probably running the minimum 5mhz 3G), T-Mo sunsetted their 2G last year and probably running the minimum 5mhz; they do plan to sunset their 3G systems in the ~2022 timeframe but it won't gain much at all.

That said, I don't know how much spectrum they need; I mean, T-Mo (from buying Sprint) has the most average low and mid-band, something like 250mhz average (Sprint had like 150mhz at 2.5ghz...) while AT&T has around 120mhz and VZW averages like 80-100mhz or so. This 100s of mhz of new spectrum in CBRS and C-Band alone is more than most of these companies had to begin with (let alone all this other spectrum the FCC wants to free up...), so I do think there'll be a point soon where they really have no use for more and more spectrum.

CBRS is kind of interesting, there's a section where it's kind of "free for all", essentially unlicensed as long as the equipment follows some kind of access rules that splits it up between all users fairly; and a "priority access license" section where it's licensed, but if the cell co or whoever licensed it but isn't using it it's free for anyone else to use. So, VZW for instance was expected to bid throughout the country, instead they bid high in some areas and not at all in quite a few others, hedging thier bet that under these PAL rules there'll be loads of open CBRS to use free of cost.

*Band 13 is not an approved band for 5G, the ETSI committee that formally approves new 5GNR bands did not meet last year due to travel restrictions, and I have no idea why they didn't just do it online.