The USA has been the biggest deployer of 5G in millimeter wave (mmWave) bands in the world, though mainly through necessity rather than choice. But its operators and their vendors are learning a great deal about how to use this challenging spectrum cost-effectively, and this is being fed into real world solutions as well as into further R&D – for instance, around the potential use of 70 GHz, which actually has better propagation characteristics for many applications than the widely used (unlicensed) 60 GHz.
And rising interest was demonstrated by the results of Auction 103, the FCC’s third consecutive mmWave spectrum sale, which were announced last month and generated total revenue of $7.6bn (a total which is split between $3bn in incentive payments for existing licensees in the bands and $4.5bn in net FCC proceeds). Last week, the agency announced the details of which operators had won licences, with Verizon once again being the most enthusiastic investor in mmWave.
Auction 103 spanned the upper 37 GHz, 39 GHz, and 47 GHz band, following on from Auction 101, in 28 GHz, and Auction 102, in 24 GHz. In the most recent process, 99.9% of available licences were snapped up. By contrast, Auction 101 raised $703m in total bids, and Auction 102 raised $2bn. Verizon, AT&T and T-Mobile were the main purchasers in both, as they were in Auction 103.
In this most recent sale, Verizon spent about $1.6bn for 4,940 licences across the country; while AT&T was close behind, buying 3,267 licences for around $1.2bn. The biggest surprise was that T-Mobile, which had been predicted to be the biggest buyer, only spent $873m, for 2,384 licences – perhaps because, at the time of bidding, it had become reasonably confident that its acquisition of Sprint, with its valuable and plentiful 2.5 GHz midband spectrum, would go ahead, making mmWave capacity less critical.
While TMO was expected to spend more, it is surprising that Sprint took part at all, given its financial challenges, its impending acquisition, and its ownership of well over 110 MHz of midband spectrum in all markets. It has frequently said it does not need mmWave because of the ideal balance of capacity and coverage which its 2.5 GHz spectrum supports for current 4G and future 5G. However, it bought only 127 licences, so perhaps these were for opportunistic reasons in selected markets.
Dish Network, which will gain spectrum and network assets to launch its 5G network once the Sprint/TMO merger is finalized, has nevertheless decided to add mmWave assets to its spectrum patchwork with a $203m investment. While Sprint is an example of an operator which has tried to consolidate as much of its mobile broadband as possible into a single band, to improve scale and simplicity of deployment and device ecosystem, Dish an extremely mixed pot of different airwaves from former mobile satellite frequencies to AWS-3, 600 MHz, mmWave and others.
This may be because it originally regarded its spectrum purchases mainly as tradeable assets – it has not built out a network of any significance, though it will not have to under FCC mandate. But it also reflects how modern and greenfield MNOs are less bothered about common bands, and more interested in aggregating diverse spectrum using software-defined networking and orchestration techniques, plus emerging wideband antennas in infrastructure and devices. This enables them to pick up any spectrum that looks affordable to create a significant combined pool.
The top 10 spenders in Auction 103 were:
|Bidder (bidding entity)||Net payment||Number of licences|
|Verizon (Straight Path Spectrum)||$1,624,101,808||4,940|
|AT&T (FiberTower Spectrum Holdings)||$1,185,734,976||3,267|
|T-Mobile (T-Mobile License)||$872,791,192||2,384|
|Columbia Capital (High Band License Co)||$306,711,619||52|
|Dish Network (Window Wireless)||$202,532,574||2,651|
|Sprint (ATI Sub)||$113,948,318||127|
|Monarch Wireless (a bidding entity)||$29,461,717||173|
|NKCN (a bidding entity)||$6,718,504||68|
Note – most operators in the USA take part in auctions through a bidding entity. Verizon’s and AT&T’s bidding entities take the names of the respective companies they acquired in 2018, to add to their mmWave assets ahead of any auctions – deals which helped them make their very early commercial 5G launches for fixed wireless access in mmWave bands.
Verizon, by contrast with TMO, had not been expected to bid heavily since it is already by far the strongest in terms of mmWave spectrum ownership. It has spent billions of dollars on licences, in the previous two auctions, and also with a string of acquisitions of companies that brought it mmWave spectrum, among other assets. Those acquisitions include Straight Path, XO Communications and Nextlink – all of them businesses which emerged from bankruptcy ashes after the USA’s broadband wireless access bubble in LMDS (28 GHz and 39 GHz) spectrum, at the start of the century, burst.
Verizon also acquired Sunshine, Highland Holdings and other smaller owners of mmWave assets, while AT&T bought FastPath.
The heavy spending by Verizon highlights its very different approach to early 5G, compared to its rivals. AT&T – and to an even bigger extent, TMO – clearly saw mmWave as a way to get into the 5G market early, but with their eyes firmly on middle and lower bands. But Verizon is trying to build a business model around long term and specific use of mmWave, and fixed wireless, to extend its fixed-line and multiplay footprint, which is far smaller than AT&T’s. Auction 103 suggests that it wants to double down on its spectrum advantage in the high bands, preventing any other operator being similarly positioned at the stage where mmWave becomes a far more mainstream and deployable foundation for 5G in future.
Overall, 28 bidders in Auction 103 won more than 99.9% of the available licences. The next auctions will be for the licensed elements of the CBRS scheme in 3.5 GHz, scheduled to begin on June 25, and the C-band auction, scheduled to begin on December 8.