Early on Thursday, AT&T said it would be the first US operator to launch a mobile 5G service, in 12 markets in "late 2018," but the plans announced so far leave some fairly significant questions unanswered.
For instance, I asked an AT&T spokeswoman what 5G devices will be available for the service. This is important because most in the industry don't expect 5G smartphones to be available until sometime in 2019.
"It will be a mobile device," AT&T says, offering no more details. The carrier is also not yet revealing what spectrum it will use for the 5G service. "More to come there, " the spokeswoman says. (See AT&T Joins 5G Marketing War, Promising 'Mobile' Launch in 2018.)
Like I said, that leaves two of the major questions about AT&T's mobile 5G plans unanswered.
The key question is the spectrum. I suspect that AT&T will use some of its existing spectrum -- possibly even some unlicensed spectrum -- and channel aggregation techniques to deliver its first 5G service.
Several AT&T execs said last year that the operator had around 60MHz of "fallow spectrum" for use on LTE. This includes AWS-3 mid-band spectrum and higher band 2.3GHz (WCS) spectrum, although AT&T has started deploying LTE in both bands. (See AT&T Expects 5G in Late 2018 or Early '19.)
This would still likely leave it with less than the recommended 100MHz channel for 5G. That, however, is a problem that will also face T-Mobile with its planned 600MHz deployment.
AT&T bought FiberTower in January 2017 for its "5G" 24GHz and 39GHz licenses. The FCC, however, hasn't yet approved the transfer of those licenses, even in its massive end-of-the year license approval document, released on December 27. (See AT&T Buys FiberTower for 5G Spectrum.)
So, frankly, if AT&T is targeting those frequencies for a mobile 5G launch in late 2018, it's running out of time, if it hasn't already.
Similarily, even if the FCC auctions off millimeter bands tomorrow, which it won't, operators would still have a hard time preparing that spectrum for use this year.
As for the 5G "mobile device" that AT&T is promising in 2018, the phrasing alone suggests that it may not be a smartphone per se. (See Qualcomm: First 5G Smartphones Coming Mid-2019.)
A tablet seems the next most likely option. Samsung is already slated to supply 5G test tablets for the Winter Olympics in Pyeongchang, so that's not a huge stretch.
What we can say for sure -- much like the first 3G and 4G launches -- is that early 5G in the US will be as much about bragging rights and marketing as it is about a revolution in wireless services. These early examples will be just like 4G, but faster -- when they're not falling back to a 4G connection, anyway.
The revolution will have to come later, and if history teaches us anything, it's that the revolution probably won't come from the carriers anyway.
— Dan Jones, Mobile Editor, Light Reading