The US regulator, the FCC, will kick off its third auction of millimeter wave (mmWave) spectrum on December 10, but the large cable operators are sitting this one out, which may prevent prices going too high.
All the national MNOs are on the list of those registered to participate in Auction 103, which will offer a massive 2,400 MHz of spectrum across the 37 GHz and 39 GHz bands, and a further 1,000 MHz in the 47 GHz band. However, Comcast, Charter and Altice USA, despite their high interest in adding mobile and small cell services to their cable networks, are not on the list.
So far, prices in the mmWave auctions have been far lower than for spectrum in lower bands, notably the 600 MHz incentive auction of 2017, which generated more than $20bn. Of course, there is more spectrum available in mmWave bands, so the per-MHz/pop price is always likely to be lower because there is no scarcity value, but at $2.7bn combined, the previous mmWave auctions delivered less in absolute terms too.
The cablecos seem to be pinning their hopes of mobile self-sufficiency – reducing their reliance on MVNO deals with Verizon or Sprint – mainly on CBRS spectrum in 3.5 GHz, either unlicensed or by bidding for licences in next year’s auction. The CBRS band would allow them to build out small cells and localized networks to augment their cable and WiFi connections, leveraging those cables for indoor and outdoor backhaul. But it would not require the large number of cells, nor the engineering complexity, of the high frequency spectrum.
Other companies looking to bid in the next auction include Dish Network, which is planning a fully virtualized 5G network in its patchwork of spectrum, assuming the Sprint/T-Mobile merger is approved, as this will result in Sprint divesting spectrum, sites and users to Dish and supporting some infrastructure sharing. Windstream, US Cellular, GCI and Columbia Capital (a venture capital firm which frequently invests in spectrum assets) are also registered.
Some small providers are also listed but are only likely to bid in their local areas. These include OptimERA, LICT Corporate and VTel.
Despite their hoped-for merger, Sprint and T-Mobile are both registered to participate – the FCC has issued a waiver to allow them to communicate during the auction, which would usually not be allowed during and FCC process.
Brian Goemmer, president of AllNet Insights & Analytics, told FierceWireless that there may be fairly low demand and prices in the upcoming sale, because of the sheer amount on offer, and because so much mmWave spectrum has “already been brought to the marketplace that hasn’t even started to be used yet”. This refers not only to the two previous auctions – of 24 GHz and 28 GHz licences – but also to 28 GHz and 39 GHz spectrum which both AT&T and Verizon have bought via acquisition over the past couple of years.
In the 24 GHz sale, AT&T and T-Mobile were the biggest investors – AT&T spent around $980m on 831 licenses covering 383 partial economic areas (PEAs), to achieve about 600 MHz of mmWave spectrum, on average, nationwide. T-Mobile bid about $800m in the 24 GHz band and $39m in 28 GHz, while Verizon spent about $505m on additional 28 GHz licences and some 24 GHz airwaves. AT&T and Verizon are currently operating on temporary licences in the 39 GHz band, while the 37 GHz and 47 GHz bands are unencumbered.
Verizon has about twice the amount of mmWave spectrum of its rivals, but all three have enough to be going on with, believes Goemmer, and may not bid high prices in Auction 103. “Even where AT&T and T-Mobile have 50% of the spectrum assets, roughly, that Verizon does, is it enough for them to offer 5G services like VR/AR, gaming, and super-fast movie downloads with the spectrum they already have? The answer is probably yes,” he said.
Most early 5G deployments in mmWave bands have looked rather similar to the broadband wireless access (BWA) roll-outs, in the same bands, that created a shortlived bubble in the USA at the start of the century – focused on fixed wireless access (FWA) in suburban areas, with outdoor base stations serving CPE on the windowsill or outside the house.
But there are indoor use cases too and Qualcomm and Nokia have both talked extensively about how mmWave spectrum could enable next generation 5G in-building services from smart factories to superfast office and home networks supporting applications like holographics.
US-based RAN and backhaul vendor, JMA, plans to build a 5G manufacturing plant in Syracuse, New York, for what it claims will be the world’s first 5G mmWave radio system specifically commercialized for indoor use. This will see JMA building on its old business in distributed antenna system (DAS) while adding its newer credentials as an open virtualized RAN vendor and base station antenna designer.
The decision to open a local factory will play well in a country which is looking to score points against China in the ‘race to 5G’, while also extending the choice of vendors for MNOs which are not able to use Chinese equipment.
Last year, JMA bought mmWave antenna start-up PHAZR, which had trialled its equipment with AT&T. Its product will be the heart of the indoor mmWave product line, which should start to be manufactured in Syracuse in early 2021, according to Greg Dial, who heads up JMA’s edge compute division, Cusp.
“We’re really on the doorstep of completing what will be the world’s first 5G millimeter wave in-building radio solution,” he said.
He explained that the mmWave product will not be merely a receiver for a signal generated outdoors, like FWA systems, but “is directly generating the 5G millimeter wave signal indoors … This is really bringing 5G millimeter wave, the last mile, to the user in a way that is not trying to do trick shots from bringing it in from outdoors.”
With the radio inside the building, “the same physical properties that make it difficult to get that millimeter wave signal indoors is actually going to work in our favor once we have the radio physically inside,” he added, citing manufacturing and AR/VR as key use cases.