Your browser is not supported. Please update it.

25 March 2016

As auction looms, 600 MHz will be less important than mmWave to USA

The US incentive auction of 600 MHz broadcast spectrum is set to start on Tuesday, though the intense attention of last year has been partly diverted to other future sources of mobile spectrum, including the millimeter wave bands, which will prove as controversial as the TV airwaves – and, in a capacity-obsessed world, more important to the future expansion of wireless services.

The impact of the 600 MHz winners and losers will be diluted by the growing attention to high frequency, ‘5G’ bands, and by the time it will take to activate the spectrum commercially. FCC chairman Tom Wheeler this week acknowledged, to the House Subcommittee on Communications and Technology, that the 600 MHz spectrum would take years to repurpose for mobile broadband. The auction itself could take months, and then the “transition will be a complex, multi-disciplinary effort that will span several years,” Wheeler said.

The operators, as Sprint has made clear, may be shifting their attention to higher frequency, capacity-oriented bands, but the 600 MHz sale still has the potential to shake up the competitive landscape, in the short term at least. The FCC released the list of applicants this week, with no fewer than 104 names (though 35 were incomplete applications, including AT&T’s). Verizon, T-Mobile, US Cellular and C Spire were all present and correct, as well as their potential nemesis, Comcast, bidding under the name CC Wireless Investment.

However, it remains to be seen how much of an impact Comcast will have – as the leading cableco, and deployer of a vast network of WiFi homespots and hotspots, it has the ability to add licensed spectrum and become a major quad play powerhouse. But it has indicated it does not intend to spend inflated sums in the auction, and could retreat to using its MVNO agreement with Verizon instead.

Two other companies whose participation could boost the prices and also shake up the market are Softbank of Japan, Sprint’s owner – which indicated recently that it might bid in its own right, even though Sprint is staying out – and Dish Network, which may seek licences either as tradeable assets or to add to its store of spectrum for a future ‘5G’ roll-out. Neither firm appears on the list of applicants under its own name, but may be operating behind a proxy – FierceWireless pointed out that one name, ParkerB.com Wireless, is registered in Denver, also home to Dish. And another Japanese operator, NTT Docomo, has registered an application, under the name Docomo Pacific, though BTIG analyst Walter Piecyk speculated, in a client note, that it would not bid on continental USA licences – and nor would another foreign applicant, America Movil.

Movil could be thinking of expanding its cross-border activities in response to AT&T’s entry into its home market of Mexico, and it already has a US-based MVNO. But given the cost involved in acquiring significant amounts of spectrum in major US markets, and building networks, it seems more likely it will confine its interest to Puerto Rico and possibly Hawaii.

Most of the cable operators have stayed out of the running. This will mark a significant difference from the auctions of a decade ago. In 2006, a group of leading cablecos (Comcast, Cox, Time Warner Cable and Bright House) joined with Sprint to form the Spectrumco consortium and acquired licences in the AWS band; in 2008, Cox acquired spectrum in the 700 MHz auction (most of these assets were later sold to Verizon).

But cable firms like Cablevision are now focusing mainly on WiFi, backed up by potential MVNO activities, to add wireless elements to their multiplay bundles. Two of the participants in the former consortium, and the defunct Pivot joint venture with Sprint, are soon to become a single firm – Charter is in the process of acquiring both Time Warner Cable and Bright House, and probably saw the auction as a distraction from that. Neither Cox nor Liberty Global have submitted applications.

Another major difference from the 2008 auction is that Google is not taking part. Back then, it helped boost the price of the only nationwide 700 MHz block (which Verizon won), and to force adoption of open access provisions, by threatening to buy the asset.

Eight years later, its focus has shifted. Its overall objectives remain the same – to increase the store of spectrum available to support new wireless broadband capacity, and to ensure this is open to a wider variety of service providers, in order to accelerate broadband availability and lower costs, thereby expanding the market for Google services and adverts. But rather than shaking up the regulations and ownership of conventional cellular bands, it is concentrating its lobbying and build-out efforts on additional unlicensed and lightly licensed bands, such as the 3.5 GHz Citizens Band and the TV white spaces, and for future 5G systems, the centimeter and millimeter wave spectrum above 6 GHz (see inset).

In this area, Google has been awarded a two-year experimental licence by the FCC, to conduct national tests of airborne and terrestrial transmitters in the 71-76 GHz and 81-86 GHz mmWave bands. The FCC says the licence is conditioned on not causing any harmful interference to other authorized trials or operations (some wireless fiber systems work in these bands), and conforming to RF safety standards (some parties had objected to the trial on the grounds of potential health risks).

The likely subject of the mmWave tests will be Project Loon, which uses both airborne and terrestrial transmitters – the name on the licence application was Astro Teller, director of Google X, the lab which houses the firm’s ‘moon shots’, including Loon. The application also refers to “airborne max altitude 75,500 ft. AGL,” which is consistent with Project Loon’s high altitude balloons, though also with drones like those of Titan Aerospace, acquired by Google/Alphabet in 2014. Providing broadband access, not just backhaul, via airborne platforms such as drones and balloons, is a key focus of Google’s R&D program to get access to underserved areas where terrestrial build-out is difficult or cost-prohibitive.

Another very active participant in the early race to colonize mmWave spectrum is T-Mobile, which owns 200 MHz in the 28 GHz and 39 GHz range courtesy of its purchase of MetroPCS in 2012. Like Verizon (which gained access to 28 GHz and 39 GHz spectrum via its recent acquisition of XO), TMO plans to conduct trials this year in high frequency bands. Its assets in these airwaves could give it a first mover advantage in 5G trials, write Wells Fargo analyst Jennifer Fritzche in a recent research note, especially as AT&T does not have any of these frequencies, and Sprint has access to them only in Alaska.

However, the bulk of the mmWave spectrum remains in the FCC’s hands – a great deal was returned to the pot after the failure of several LMDS fixed wireless businesses earlier in the decade, so it should not be hard for players to gain temporary licences to conduct 5G tests and further the agency’s ambition of putting the US in the 5G driving seat.

TMO’s CTO Neville Ray said, on the operator’s fourth quarter earnings call last month: “We’ll make sure that there’s nothing in the 5G space that we don’t deliver to our customers on or before our competition does … We’ve already got significant spectrum holdings in what we believe will be declared 5G spectrum allocations. That doesn’t mean anybody has the spectrum that they need for a full 5G vision, but we’re out in front and the other guys for their trials are filing STAs [special temporary authorities], they’re having to get test licences. We already have a big swath of spectrum across many parts of the country that we can use for 5G tests and trials.”

Of course, there are many decisions to be made on how these high frequency bands will be regulated and shared – between different mobile providers, and between the wireless and satellite sectors. Both these issues are already attracting significant controversy.

TMO, for instance, is opposed to one of the FCC’s ideas – to extend the flexible Spectrum Access System (SAS), introduced for the shared portions of the 3.5 GHz band in the US, to mmWave in future in order to avoid interference (for instance, with satellite users). John Hunter, TMO’s director of spectrum policy, argued at an FCC workshop, Spectrum Frontiers, held earlier this month, that such a system might not be technically workable in mmWave spectrum.

“I think we need to take a step back and really think about what the SAS was intended for [military radar operations],” Hunter said. He said sensing mechanisms will be far more challenging in 37 GHz than in lower bands, and the system will not be dealing with unchanging “known locations” as they are with radar. AT&T agrees, and wants the 37 GHz and 39 GHz bands to be amalgamated into one major band, to be licensed on a flexible use basis.

It seems increasingly likely that the hybrid licensing approach of 3.5 GHz will be dropped as a proposal for mmWave, given the weight of feeling against it. Wheeler told this week’s sub-committee hearing: “We’re approaching 5G as we have with previous generations of wireless by adopting a flexible use policy and assuring that spectrum is available to be deployed when the private sector has arrived at the requisite technical standards and network architectures.”

Vendors are also calling for a common, flexible licensing approach between 37 GHz and 40 GHz, rather than a hybrid scheme – Ericsson, Intel and Nokia recently met with the FCC to make that case, while Samsung echoed the same opinions when it took part in the workshop, and also called for current LMDS fixed wireless licensees to be allowed to support mobility in their spectrum. Qualcomm wants the FCC to use the limits and test procedures set out in the IEEE and ICNIRP standards for portable device operations in the mmWave bands.

Intel, Nokia, Samsung, Qualcomm and Ericsson have also been demonstrating technology for mmWave bands to the FCC, including multi-antenna array, steerable beamforming, massive MIMO, ultra-dense networks and other technologies.

Samsung showed off a 3.7Gbps data link using an 800 MHz channel in the 28 GHz band, using prototype indoor equipment equipped with TDD, high order modulation and MIMO with beam selection.
Ericsson’s main demonstration was remote control of a Volvo vehicle in Dallas from a connected simulator placed in the workshop’s location in the FCC’s Washington DC headquarters. Nokia showed off an experimental 5G system in the 73 GHz band, boasting latency below one millisecond and multiuser acquisition and tracking, while Qualcomm showed live streaming over its own 5G prototype. Intel demonstrated its 5G Mobile Trial Platform in 60 GHz, Mobile Edge mmWave backhaul and access links for small cells, its pre-5G anchor booster concept for mmWave, and the emerging NB-IoT technology operating in high frequencies – all also seen at Mobile World Congress.

But at the workshop, FCC Commissioner Michael O’Rielly also called for concrete proposals about site and network planning issues, not just technology. “Even if we were to have infinite spectrum resources, without substantial network deployment, the American people will not benefit from 5G,” he said.

He told attendees: “The densification of networks will result in an increase in facilities serving as access points and providing backhaul, which leads me to a topic I discuss often – the need to reduce burdens on infrastructure siting.  To realize the promise of 5G, companies will need to expeditiously deploy equipment in a cost-effective manner.  To facilitate network deployment, the Commission must fulfill its commitment to expand upon the historic and environmental process relief provided to small cell and DAS installations.  But even this action, which will be tremendously helpful, is not a cure-all.  I continue to hear complaints about permitting problems, excessive fees and de facto moratoria.  Site approvals in rights-of-way, which are especially important for small cell systems, appear to be particularly problematic.  I hope that panelists will take some time to discuss ways to ease the burdens of facilities siting.”