Your browser is not supported. Please update it.

20 March 2020

Despite two auctions, US MNOs still have a long wait for midband 5G

The US operators remain desperately short of the midband spectrum which is the key engine of 5G roll-out for just about all their counterparts outside North America. During this year, there will be two auctions that will start to address the shortage, though that still leaves them with over a year to extend 5G services without access to midrange airwaves (with the exception of T-Mobile USA, once its acquisition of Sprint, and its 2.5 GHz spectrum, is finalized, probably around April).

The first midband auction will be of the priority access licences (PALs) in the 3.5 GHz (3550-3700 MHz) CBRS band. These are not ideal at all from MNOs’ point of view because they are local and short term compared to the usual mobile licences. Auction 105 will offer seven PALs in each county-based license area, and each PAL will consist of a 10 MHz unpaired channel within the 3550-3650 MHz band.

However, they will provide valuable assets to deploy services that require significant capacity but only need local reach, such as fixed wireless access or industrial networks. And they are expected, because of their limits in time and space, to be sufficiently affordable for new entrants, such as small cell neutral hosts or enterprise providers, to enter the market (though that will depend on the level of competition of course).

Some operators are looking to combine PAL spectrum with the general access element of the three-tiered CBRS scheme to expand capacity in a flexible way, though that will not apply to 5G until standards for 5G New Radio in shared spectrum are completed, in Release 16.

Even more contentious, in the process of defining rules, than CBRS PAL, has been the C-band spectrum, a little higher up than CBRS in 3.7-4.2 GHz. Combined, these two middle bands are the main area of the spectrum powering 5G worldwide, and in most countries have either been vacant, or relatively easily repurposed from fixed wireless, other microwave or public sector applications. But in the USA, the presence of military radar in CBRS, and satellite in C-band, have forced a more complex approach – innovative yes, but with considerable uncertainties about how challenging it will be to deploy the spectrum cost-effectively, to build a broad device ecosystem, and to coexist with incumbents.

The C-band auction is scheduled, provisionally, to start in December. Verizon has its eyes on this, as well as increasing its millimeter wave pot (see separate item). The operator’s CFO Matt Ellis told investors recently: “As we look at C-band, it brings forward some use cases that could increase the size of the 5G opportunity for us.”

Given the high capacity nature of Verizon’s 5G business model, which is also charged with delivering fixed broadband over a large footprint, the telco will be particularly pleased at the FCC decision not to impose caps on the amount of C-band airwaves one MNO could buy. The auction will only be subject to the usual FCC spectrum screen, which enables the agency to investigate, and potentially block, a transaction which would give one operator more than one-third of the total spectrum in a given market (blocks on purchases under this system have been extremely rare).

Verizon is also interested in using both unlicensed and licensed CBRS spectrum, but believes this will fulfil a very different purpose from C-band. “We’re thinking about both of them but in very different ways,” Ellis said. “If you start with CBRS … it’s something that certainly we see as being complementary within our LTE network experience initially and then eventually it could be part of 5G as well. We certainly will look at it from a 4G capacity standpoint.”

As with 600 MHz, the FCC needs to offer incentives for the incumbents to vacate the band, and in the case of C-band – given the urgent nature of the need for midband 5G airwaves – to vacate it quickly. The current occupants – SES, Intelsat, Telesat and Eutelsat – use the band to support content companies such as Disney and NPR, but they will now need to confine those activities to the upper portion of the band. The FCC is allocating the spectrum in 3.7-4.0 GHz for mobile use, with 280MHz (3.7GHz-3.98GHz) to be auctioned and the other 20 MHz kept as a guard band. The satellite users will move into the upper 200 MHz (4.0-4.2 GHz).

The FCC has set a schedule to transition 280 MHz of the C-Band for flexible use no later than December 5 2025 but it wants at least the first 100 MHz in the top 46 markets to be available for 5G far sooner. It is offering $9.7bn in incentive payments for the current users to update their satellite equipment quickly in order to free up the spectrum. In addition, C-band auction winners can negotiate with incumbents to persuade them to vacate the band even more quickly. These incentives could, the FCC hopes, enable initial deployments as early as the first quarter of 2021 in some markets.

Satellite providers will be eligible for accelerated relocation payments of $9.7bn “if they commit to, and succeed in, clearing the spectrum early”. To be eligible for Phase I of the payments, the operators must clear 120 MHz in 46 Partial Economic Areas by December 5 2021, and for Phase 2, must clear the remaining 180 MHz of spectrum by December 5 2023.

While Verizon appears to have shown its hand in terms of its eagerness for C-band spectrum, AT&T and T-Mobile are playing it cool. The latter will be getting its valuable dowry of 2.5 GHz spectrum, only some of that used for 4G, from Sprint; while AT&T’s CFO, John Stephens, told an investor conference: “We’ll certainly have an interest in the C-band auction, but quite frankly it’s not something that’s essential to us on an immediate basis because we have this significant amount of spectrum already owned and already placed in service. We don’t have the immediate need for that [C-Band] spectrum or putting new spectrum into service in the same light as some of our competitors.” These assets include WCS, AWS and FirstNet spectrum.

During a painful process of getting agreement between the satellite and terrestrial industries over C-band, the former group initially pushed for a private sale of the spectrum, which the FCC eventually rejected. The current plan involves a public auction to begin on December 8, and the incumbents must hit a number of clearing targets in order to be eligible for funds. It is not clear that they will all support this timetable. Intelsat and others are still arguing for heftier incentive payments, or for more of the sums to be paid in Phase 1. Intelsat which has a $14bn debt mountain, claims to have invested $36bn in building its C-band business over the past 40 years.

Commissioner Jessica Rosenworcel, who voted against the current plan, warned Congress recently: “We are headed fast to litigation. We’re going to the courts. It’s going to slow things down.” She and fellow Democrat Commissioner Geoffrey Starks fear that the FCC is not legally authorized to go down the current path, and that a Congress-approved framework would be safer. FCC chairman Ajit Pai argues that would take too long.

“In the political reality that we are in, the notion that we should sit around and wait for Congress to come up with complex legislation on a topic that divides virtually every sector of the industry is crazy,” he said. “I for one am not going to forfeit American leadership in 5G on the hope that political actors from across the spectrum with differing incentives can come together.” Congress could still propose its own alternative legislation, but analysts think this is unlikely given the depth of disagreements. Many lawmakers’ opposition to Pai’s plan centers on the size of the incentive payments rather than the principle of repurposing the spectrum.

The auction will have two categories of spectrum blocks for each region – five 20 MHz blocks in the lower 100 MHz (3.7-3.8 GHz) and nine 20 MHz blocks in the remaining 180 MHz (3.8-3.98 GHz). Bidding credit caps of $25m for small businesses and $10m for rural service providers are proposed, as well as a $10m cap on the overall amount of bidding credits that a small business bidder may apply to win licenses in smaller markets.