Three main MNOs win in US mmWave auctions but still lack midband assets

The FCC has announced the winners of the two millimeter wave auctions held so far this year, in 28 GHz and 24 GHz. Unsurprisingly, AT&T, Verizon and T-Mobile were the big gainers, but while this boost the capacity arsenal for all three, they all lack significant holdings in the far more attractive midrange bands, which are the main focus of early 5G roll-out in most other parts of the world. To compensate, Verizon is starting to make use of the 3.5 GHz CBRS spectrum, which is becoming available this year in a multi-tiered system of licensed and shared access.

The midband airwaves support high capacity hotspot deployment without the challenges of mmWave – the latter spectrum needs complex beamforming to get round the range limitations in these high frequencies, and there are issues of cost and power consumption which may restrict implementation in devices.

But with the USA being short of midband airwaves for cellular (with the notable exception of Sprint’s plentiful 2.5 GHz reserves), AT&T and Verizon have been forced to use mmWave spectrum in order to achieve their aggressive timescales for their first 5G services. Both acquired 28 GHz and 39 GHz assets which had originally been allocated for fixed broadband wireless, in order to gain a headstart; by contrast, TMO was the biggest investor in the 600 MHz auction last year, and plans to kick off its 5G deployment with a wide coverage layer. However, while its sub-GHz bands are good for cost-effective coverage, they have limited bandwidth, and so the 5G speeds and user experience will not be hugely different from those of 4G.

Now all three operators have topped up their mmWave reserves via Auctions 101 and 102, which generated about $2.7bn in total. The FCC had not previously announced the winning bidders but has done so now.

AT&T led the way with total bids of $982m in the 24 GHz auction, buying 49% of all the available licenses (831 of them), while Verizon led the 28 GHz sale, paying $506m for 72% of the available licenses, numbering over 1,000. These will fit well with Verizon’s existing mmWave holdings, which are heavy on 28 GHz, while AT&T has more in 39 GHz. That would explain why Verizon was a minor player in 24 GHz. According to Allnet Insights, Verizon now has 65% of the USA’s 28 GHz spectrum, though 25% remains in the FCC’s hands; in 24 GHz, TMO now has 48% and AT&T 36%.

T-Mobile was the second bidder in the 24 GHz auction with $803m in total bids, or 40% of the available licenses; it also spent $39m in the 28 GHz sale, though it was in third place behind regional player US Cellular, which spent $129m in bids. US Cellular spent a similar amount on 24 GHz spectrum, and was in fourth place in the bidding after the big three telcos.

Top 10 bidders in the two auctions:

One important aspect of the announcements is that the operators will be free to engage in M&A talks again within two weeks, though most are likely to wait for the outcome of the review of T-Mobile’s proposal to acquire Sprint. Negotiating cannot take place in the period around an auction but will now be allowed until the run-up to the next mmWave auction, which is scheduled to start on December 10.

FCC chairman, Ajit Pai, said: “The successful conclusion of our nation’s first two high band flexible, mobile-use spectrum auction is a critical step. By making more spectrum available, we’ll ensure that American consumers reap the substantial benefits that 5G innovation will bring and will extend US leadership in 5G too.”

In the midbands, Verizon has started deploying antennas to support CBRS, according to LightReading sources. The carrier has also been pushing its handset suppliers to support the band. Of course, the USA’s 3.5 GHz is a significant compromise, as far as MNOs are concerned, compared to the band’s status as the leadership 5G band in other markets including China. Because it is a federal band in the USA, it has been opened up under a shared system – federal users gain top priority access, followed by owners of licences (to be auctioned next year) and then by general access users.

If operators choose to focus mainly on general access, they will have to deploy LTE for the next few years at least, since 5G standards for shared spectrum have not been finalized. If they buy licenses, they can deploy 5G, but these franchises cover limited geographies and have only three years in duration – a far cry from the MNOs’ usual large territories and 10-year timescales, though the licenses are less flexible and localized than many campaigners had wanted. Those parties had argued that a series of short term or on-demand licences in limited geographies would have encouraged new deployers and services, especially for industrial applications.

But beggars cannot be choosers in the US midband spectrum and it seems Verizon will at least stake a position in CBRS in the unlicensed area, since its trials are coming well before the auctions are likely to take place. Verizon is also bolstering its LTE capacity, while keeping spectrum costs low, by deploying LTE-LAA (Licensed Assisted Access) to provide supplemental downlink for its licensed spectrum 4G networks, using the unlicensed 5 GHz band.

In CBRS, the telco has run trials with Federated Wireless, Google, Nokia and Qualcomm, and has said it will use the spectrum for outdoor and indoor small cells. But its commercial plans seem closer than previously expected, given that it is already deploying antennas in its network and selling several CBRS-capable handsets including Google’s Pixel 3 and Samsung’s Galaxy S10.

The next milestone will be when the FCC and NTIA give final approval for commercial deployments in CBRS, which is expected within a few weeks.