As of Monday (December 17) the FCC’s 28 GHz auction had reached $688.7m in total bids after 74 rounds of bidding. There will be a break in the process from 1pm (Eastern Time) on Friday December 21 until 10am on January 3 2019. There are now 2,939 licences in the 28 GHz band with provisional winning bids and 133 left on the table.
That auction is clearly reaching its conclusion, and the probable final prices will reflect changes in the perception of high frequency spectrum. Prices per MHz/pop will be low because of the high capacity of this band – in the upcoming 24 GHz auction, the opening prices have been set a full 188 times lower than those for the 600 MHz incentive auction. In a report issued earlier this year, Spectrum Financial Partners wrote: “The FCC plans to begin bidding for all 700 MHz of the 24 GHz band at $438m. This compares with opening bids for the 600 MHz auction ($1,174m for each 10 MHz block) in the ratio of 117.4/(438/700)= 187.6 or about 188 times higher opening prices per MHz for 600 MHz spectrum than for 24 GHz spectrum.”
However, the actual sums invested in mmWave airwaves are rising. Even in the brief broadband wireless bubble which inflated around LMDS spectrum in 1998-2000, prices were only fractions of a dollar per MHz/pop, and in other countries where licences were awarded in similar bands, prices were even lower.
However, those unloved LMDS frequencies (28 GHz and 39 GHz) are newly attractive because of 5G, and Verizon and AT&T pushed up the value attached to such assets when they bought up owners of those existing licences. Verizon outbid AT&T to pay $3.1bn for Straight Path, the owner of substantial 28 GHz and 39 GHz assets, which it had acquired from bankrupt LMDS operator Winstar. Meanwhile, AT&T secured some 39 GHz licences with its recent purchase of Fibertower while Verizon gained the rights to 28 GHz spectrum as part of its acquisition of XO Communications. Analyst Craig Moffett of MoffettNathanson estimated that the Straight Path bidding war pushed the value of millimeter wave spectrum from about $0.009 per MHz/pop, to $0.017 per MHz/pop.
Nevertheless, operators will still be paying substantially less than 10 cents per MHz/pop for mmWave licences, compared to just under one dollar for 600 MHz (up to $1.40 in the top markets), and way below one of the most expensive auctions ever, the European 3G sales in 2.1 GHz (average more than $5 per MHz/pop).
Given the emphasis on 5G use cases which require very high data rates and low latency, and so lend themselves to densification, this could prove to be a good investment, as long as operators remember that – unless they have a compelling fixed wireless case, like Verizon, and even then, if they can operate in smaller cells at high power – the return on investment may take years to materialize. Technical and market changes are needed before the mmWave bands finally come into their own (see following item).
The FCC, however, unlike most regulators, is determined to lay the groundwork now and take a world-leading role in allocating high frequency spectrum for 5G, thus helping to drive further R&D on the technical and device side, and to build an ecosystem.
Earlier this year, the agency announced a plan to follow the 28 GHz auction immediately with one in 24 GHz. Now, it also plans to sell licences in 37 GHz, 39 GHz and 47 GHz next year – a massive total of 3,400 MHz on top of the capacity involved in the current sale.
There are two areas of risk in this agenda. One is that the FCC’s allocations do not align with the bands identified, at World Radio Conference 2015 (see inset), as mmWave study items for the International Telecommunications Union (ITU). The US operators may present the WRC-19 conference with a fait accompli in commercialized bands, notably 28 GHz, but they also risk have some of their airwaves classified only as regional allocations, reducing the opportunity for a globally harmonized arrangement to boost large-scale device ecosystems and international roaming.
The other is that the FCC has moved away from its initial caution – when it set out its Spectrum Frontiers program in 2015 – about clashing with the satellite industry. Satellite providers are incumbent in some of the mmWave bands which the FCC is eyeing for 5G, and so these auction consultations are unlikely to proceed as (relatively) smoothly as the 28 GHz process.
In 2015, the FCC was careful – like the WRC-15 decisions on mmWave candidates for 5G – to avoid outlining plans which would create direct conflicts with satellite. Indeed, two of the commissioners complained that the FCC should have included more bands, including 24 GHz, 32 GHz, 42 GHz, 70 GHz and 80 GHz, and had instead left “thousands of MHz of spectrum on the cutting room floor”.
But now the high frequency spectrum is in such high demand that it will be impossible to please both industries. The operators’ representative body, CTIA, has already called on the FCC to allocate almost all the spectrum 24 GHz to 50 GHz to mobile broadband, and support shared access in the 70 GHz band. It proposed squeezing the satellite players into part of the 50 GHz band, on the basis that (it claimed) their current frequencies were underused, and letting them keep the 81-86 GHz band for uplink capacity.
So the next wave of mmWave auctions is sure to provoke a showdown with the satellite industry, despite FCC chair Ajit Pai’s insistence that the process will run like the incentive auction in 600 MHz, in which broadcasters were incentivized, but not forced, to surrender airwaves for mobile broadband usage.
Far from being a model of how to reallocate spectrum among different industries in a constructive way, the incentive auction was delayed several times because of warring interest groups and the sheer complexity of its process. It took place in two stages – a reverse auction in which incumbent terrestrial broadcasters set the price at which they were willing to give up 600 MHz spectrum and move to other bands. The airwaves which were relinquished were placed into the forward auction, but could only be sold if they met the reserve price set by their current owners.
This time, the FCC says satellite incumbents will be offered “incentive payments” to give up their spectrum, in a bid to hasten the process; and they will be allowed to re-bid for the new licences. Indeed, they will be given vouchers, equivalent in value to their existing holdings, which can be used as credit towards these bids.
If a licensee refuses to join in the scheme, it will have its licence modified to align with the new-look band plan and service areas.
“This incentive auction will be different from the broadcast incentive auction that Congress authorized years ago, but it’ll have the same worthy goal: clearing or repacking existing licensees to make spectrum as useful as possible, boosting competition and benefiting consumers,” said Pai, in a statement.
The Satellite Industry Association (SIA) is not looking positive about this optimistic agenda. The sector is already waging war with the mobile industry, in many countries, over the encroachment of 5G into the upper portions of the C-Band (3.4-4.2 GHz). In the USA, the SIA responded to the CTIA’s proposal in an uncompromising way, saying such a plan would represent “an unjustified and unsustainable abandonment of the Commission’s long-standing commitment to the principle of technology neutrality”.
If the C-Band example is anything to go by, satellite industry opposition could lead to a prolonged and acrimonious consultation period, which could derail Pai’s plans to auction the full 3,400 MHz early in 2019, and even delay the sale by years. The government plans to set up a Spectrum Strategy Task Force, which will include representatives of the National Space Council, among others, and which will consult with the FCC on how best to implement the spectrum strategy.
The FCC’s chosen mmWave bands do not align entirely with the ITU’s:
Bands chosen at the ITU’s World Radio Conference 2015 as bands under consideration for 5G:
- Those which have allocations to the Mobile Service on a primary basis: 26 GHz (24.25-27.5 GHz), 37 GHz (37-40.5 GHz), 42 GHz (42.5-43.5 GHz), 45 GHz (45.5-47 GHz), 48 GHz (47.2-50.2 GHz), 50 GHz (50.4-52.6 GHz), 70 GHz (66-76 GHz), 80 GHz (81-86 GHz).
- Those that may require additional allocations to the Mobile Service on a primary basis: 32 GHz (31.8-33.4 GHz), 40 GHz (40.5-42.5 GHz), 47 GHz (47-47.2 GHz).
Bands chosen as candidates for 5G by the FCC:
- The specific bands that will be studied for 5G services are 28 GHz (27.5 to 28.35 GHz), 37 GHz (37 to 38.6 GHz), 39 GHz (38.6 to 40 GHz), and 64-71 GHz.
The first two US mmWave auctions, in 28 GHz and 24 GHz, will, together, offer 1.55 GHz of spectrum across about 6,000 geographic licence regions. Two 425 MHz blocks in the 28 GHz band are offered on a county basis, while seven 100 MHz blocks in the 24 GHz band will be allocated by Partial Economic Area.
In fact, the 28 GHz licences only cover 23.7% of the US population (and 61% of the geographic area, in 47% of counties), according to Spectrum Financial Partners (though AT&T and Verizon, and some smaller firms, already own further 28 GHz spectrum which has been commercially available since LMDS licences were sold late in the last century).