Pages

8 June 2019

We’re Running Out of Spectrum for Both New and Old Technologies

By JANE C. HU

5G is coming, and with it, the promise of faster internet that will support innovations like streaming virtual reality, self-driving cars, and smart home and city infrastructure. But before it rolls out to the masses, critical infrastructure details will need to be ironed out—for instance, what radio frequency bands telecommunication companies can use for this new technology. That’s a decidedly unsexy detail, and one most people don’t think about much. We want our radio and TV broadcasts and cellphone reception to just work. Who cares how? But now might be a good time to start caring. New technologies will require frequency bandwidth, leading to clashes between new, shiny innovations and older technologies we still need but take for granted.

Right now, there’s a conflict brewing over a specific spectrum band. Radio, TV, cellphones, satellites, submarines, Wi-Fi, and any other device that can send or receive messages do so using a designated portion of the electromagnetic spectrum to prevent interference between different types of devices. Currently, weather forecasting uses the band between 23.6 and 24 GHz, but in recent years, telecommunications companies have been eyeing the band around 24 GHz for 5G use. NOAA and NASA have raised concerns to the Federal Communications Commission that 5G’s use of nearby bands could affect their ability to keep tabs on hurricanes. Neil Jacobs, the administrator of the National Oceanic and Atmospheric Administration, told the House Science Committee earlier this month that interference from neighboring bands could result in a 77 percent data loss. Jacobs says that would have both immediate effects, like limiting our ability to predict and track developing storms, and cascading ones, like the potential for government agencies like NOAA to halt the launch of new satellites, since many of them would no longer be effective. But the FCC appears to be pushing on. In April, it auctioned off about 2,900 licenses to use the band between 24.25 and 25.25 GHz, and bidding officially ended Tuesday.

In a recent public statement, the Aerospace Industries Association wrote that its members “are excited about the promise of new technologies like 5G, but believe it is critical that the proper testing takes place prior to any FCC auction, to protect existing users of both the specific and adjacent spectrum bands.” Sens. Ron Wyden of Oregon and Maria Cantwell of Washington also wrote a joint letter to FCC Chairman Ajit Pai, urging the FCC to listen to NASA, NOAA, and the American Meteorological Society and halt commercial 24 GHz operations until potential interference effects are better understood and ameliorated.
Like land, frequency bands are finite spaces.

These conflicts will only happen more often in the future. Bandwidth is a finite resource, our demand for its use is only increasing, and much of the spectrum has already been apportioned. Spectrums currently in use for broadcasting or telecommunications were selected because they’re optimal for their intended purposes, and though some new tech may try to make use of less popular bands, inevitably, there will be competition. Basically, most of the good spectrum bands are taken, yet demand for them will continue to grow. “The iron laws of physics being what they are, we are simply not making more,” says FCC Commissioner Jessica Rosenworcel. “So the challenge is to use spectrum we have more efficiently. That also means we have to rethink how we allocate our airwaves.”

That allocation is currently up to the National Telecommunications and Information Administration, which manages government frequency use, and the FCC, which has granted licenses for commercial frequency use since 1994. Before that, the government issued licenses via applications (too cumbersome) and experimented with a lottery system (could allow winners to “flip” their licenses). Eventually, it settled on an auction system, which is “intended to award the licenses to those who will use them most effectively.” The idea is that companies’ auction bids reflect their real interest in actually using the license, rather than sitting on it as an investment to sell elsewhere later on.

Rosenworcel says the FCC has issued more than 44,000 licenses and raised “billions and billions of dollars” for the U.S. Treasury. Between 1994 and 2000, the FCC’s 24 spectrum auctions raised $20 billion, according to testimony from an adviser to the FCC. Auction 73, a contentious auction in 2008 for use of the 700 MHz radio frequency band, raised another $20 billion on its own. And there are plans to auction off licenses for three more midrange bandwidths (37, 39, and 47 GHz) by the end of the year.

But those billions mean it can get pretty expensive to play the auction game at all. The 24 GHz band auction raised nearly $2 billion with just 2,904 licenses; that averages out to $684,573 per license. The bidding system is anonymous, a feature “designed to give rural service providers and small businesses the opportunity to bid aggressively against larger carriers and included 15-25% discounts for qualified small bidders,” according to the FCC’s website, but even with help, a startup is going to have a hard time competing with AT&T and Verizon. And though the FCC hopes auctions dissuade “flipping” licenses, the system does still allow telecom giants with deep pockets to apply for licenses in certain bands and sit on them as an investment. Like land, frequency bands are finite spaces, and unforeseen future circumstances could drive up demand for that space, netting huge profits. (Indeed, some companies are counting on this.)

While issues like net neutrality and digital security have received a lot of public attention (and rightfully so!), bandwidth allocation hasn’t yet broken into the mainstream. But the current concerns around the 24 GHz band show what can happen if it’s up to the government to allocate access: It’s possible that commercial interests (i.e., an avenue for the government to bring in billions) can interfere with technology for the public good, like weather forecasting. Some legal scholars have argued that the radiofrequency spectrum should perhaps be a commons instead, where the government’s role is not to allocate rights but to ensure that users follow the rules. Whether that’s actually a tenable system, let alone better than the current one, is still up for debate.

But one recent FCC bandwidth reallocation could be a glimpse into a more “commons”-based use of the spectrum. In 2016, the FCC voted to reallocate a spectrum band around 3.5 GHz that was previously reserved for the U.S. Navy’s radar systems and reclassified its channels for use not only by the Navy but also commercial bidders and unlicensed users, along with a system to manage these different sources of traffic. “The result is on-demand sharing in the 3.5 GHz band—a major milestone in spectrum sharing and management,” says Rosenworcel. This creates the potential for unlicensed users to take advantage of this resource. They could, in theory, develop new technologies using this band without having to pony up the fees to bid in auction. It’s a new way smaller companies could get in at the ground level. It’s also a way for the FCC to figure out how best to handle multiple sources of traffic.

But first, they’ll need to address this conflict between 5G developers and weather forecasting at 24 GHz. The example the U.S. sets could have far-reaching consequences. In an internal memo, the Navy worries that other countries will follow suit in developing the 24 GHz band for 5G, and that interference to weather technology will “eventually be worldwide.” As the U.S. sprints toward “winning the 5G race,” perhaps we should ask ourselves: At what cost?

No comments:

Post a Comment