Last night, Axios broke the news that officials in the Trump administration were considering a fundamental rethinking of the US telecom industry. A proposal presented by a senior National Security Council official laid out a plan for aggressive federal intervention in 5G wireless technologies, with the federal government laying its own infrastructure to be leased to carriers. The White House is already distancing itself from the proposal, along with the FCC, but it’s noteworthy that the idea was circulating at all. It’s a bold, striking proposal that cuts against the last century of US telecom policy — and it would be wildly destructive if the White House ever did follow through on it.
In broad strokes, there’s a lot to like about the idea of a national telecom. If you’re concerned about the power telecoms have over the internet (which we are), a publicly owned telecom could be a huge force for the good, particularly if it brought connectivity to underserved rural areas. There’s never been a nationalized telecom in the United States, but lots of countries have one. As Trump looks to pivot to a populist infrastructure push, maybe a bold push for federal 5G network is just what we need.
Unfortunately, what’s laid out in the document published by Axios is very far from that idea. The main thrust behind the published proposal seems to be a fear of China, and a desire to keep them away from our telecom infrastructure. There’s a lot of hand-waving about security and the AI race, most of which doesn’t make any sense. A national wireless network would make no difference in the “AI arms race” (as the proposal calls it), and while there are many, many cybersecurity problems in the world, there’s no evidence that Huawei’s telecom hardware contributes to them in the slightest. The document makes almost no reference to the existing 5G process, which has been going on for years, and generally abandons the private industry standard system that’s guided us for the last 30 years of wireless infrastructure.
— Jessica Rosenworcel (@JRosenworcel) January 29, 2018
The key thing to understand is that 5G is not actually a network. It’s a set of standards, a broad technical specification necessary to ensure the wireless industry is working with roughly the same tools from company to company over the next 10 years. An entire generation of technologies has become available in the last decade: primarily multi-in / multi-out antenna arrays, millimeter wave, and new carrier aggregation systems. To work, those technologies need a single standard shared between device manufacturers, networks, and equipment manufacturers. To make that happen, companies have meetings and start to draw up tentative plans, creating a space they can all work within. When we talk about 5G, that’s what we’re talking about — a general consensus on how the industry will develop these technologies.
If you think of 5G as an agreement rather than a set of equipment, then it’s nearly finished. The first full spec (the non-standalone 5G NR specification) has been out for just over a month, and the standalone spec is due to be finished in June. There won’t be 5G phones or 5G cell towers for a few years, but everyone largely agrees on what they’ll look like when they get here. We know roughly what spectrum will be allocated, and what antenna technology will be supported.
Like any agreement, not everyone agrees in exactly the same way. AT&T’s approach to 5G is different from Verizon’s, just like Apple’s approach to building a 5G-capable smartphone is different from Samsung’s. That’s why the whole process is so politically delicate. The spec is forgiving, but only to a point, giving players enough room to operate while reining them in from anything that might break interoperability. If someone breaks away too far from the consensus, they end up isolated.
The race to 5G will be won by getting the government out of the way, reforming our infrastructure deployment rules, and freeing up the private sector to invest and innovate.
My statement –> pic.twitter.com/TLXPjyI8mY
— Brendan Carr (@BrendanCarrFCC) January 29, 2018
If President Trump ever followed through on the idea to build a totally standalone network, his network would end up as one of those pariahs. What’s described in the paper simply isn’t 5G: it proposes a different spectrum, different equipment, and, maybe most difficult of all, almost no overlap with the corporations that currently build and operate networks. (The US government also doesn’t own any 4G infrastructure to build on top of, making the whole thing wildly more expensive, but we’ll put that aside for now.) It would make life very difficult for carriers like AT&T and Verizon, but even putting that aside, building the network would be a real problem. The US government can decide to build its own cell towers, but it still needs base stations. There are only five companies that make those base stations, and they’re all from either Scandinavia or East Asia. Even if the US government made its own base stations, why should device manufacturers build phones for the Trump network rather than the pre-Trump consensus?
In theory, the FCC can simply refuse to allocate spectrum for the pre-existing 5G plans, leaving carriers with no choice but to adopt the federal network. But if the FCC did block existing plans for 5G, the rest of the world would keep moving forward on a unified protocol. Huawei and ZTE would still have plenty of countries to sell base stations to, and device manufacturers would be happy to make phones for them. The technical consensus, embodied in the existing specs, would allow for progress where it was still possible. A universal 5G network would happen everywhere in the world except the United States.
FCC chairman Ajit Pai surely knows this, which is why he treated the breakaway 5G proposal like poison. “Any federal effort to construct a nationalized 5G network would be a costly and counterproductive distraction from the policies we need to help the United States win the 5G future,” Pai said in a statement earlier today.
Supporting domestically built 5G infrastructure isn’t a bad idea, and you can imagine a president gently nudging the development process toward local manufacturers and nationalized infrastructure without stopping it cold. There are lots of smarter ways to do that, whether it’s tax credits, subsidies, or direct diplomacy with the standards bodies themselves.
The floated proposal does none of these things: there’s no clear upside, no foreseeable scenario in which this proposal results in more secure networks or better coverage, balanced against the very real chance of setting US networks back by a decade. It’s a half-baked idea with no technical foundation, an exercise in senseless posturing. The best we can hope for is that it stays that way.