Some old-school telecom folks will admit to being highly skeptical of millimeter wave spectrum. But with two upcoming mmWave auctions planned for this year and a bevy of mmWave spectrum swapping hands in the secondary markets, it’s clearly taken on a new life.
One of its strong points is the sheer amount that’s available. The FCC voted unanimously in July 2016 on the historic Spectrum Frontiers plan to free up vast amounts of spectrum for 5G. The move effectively quadrupled the amount of radio bandwidth ever made available to the mobile industry. The U.S. made a name for itself as a leader in mmWave and it was off to the races.
One of the pioneers in millimeter wave research, Ted Rappaport, founding director of NYU Wireless, called the 2016 FCC vote an "historic moment, a turning point, as the Renaissance of wireless begins." The professor and his students conducted ground-breaking research into the spectrum that was once deemed undesirable, proving that 5G mmWave cellular could work.
In 2017, the FCC took further steps (PDF) to make more millimeter wave spectrum available for terrestrial wireless use, noting that while mmWave bands feature short transmission paths and high propagation losses, those features can be useful in developing high capacity networks because cells can be placed close to each other without causing interference to each other. Plus, where longer paths are desired, the extremely short wavelengths of mmWave signals make it feasible for very small antennas to concentrate signals into highly focused beams with enough gain to overcome propagation losses. The short wavelengths of mmWave signals also make it possible to build dynamic beam-forming antennas that are small enough to fit into handsets.
Although millimeter wave spectrum had been around—some of its early use cases are for microwave backhaul and fixed wireless access—it wasn’t seen as worthwhile spectrum for commercial mobile uses until more recently. The 28 GHz and other high bands were intended for Local Multipoint Distribution Service (LMDS) use in the late 1990s, but those businesses failed to gain much traction and the licenses mostly laid fallow.
In fact, in large part because those LMDS businesses didn’t transpire, the licenses came available again on the secondary market. Verizon last year won a bidding war with AT&T over the licenses held by Straight Path, and AT&T ended up acquiring licenses that were held by FiberTower.
mmWave in motion
The 28 GHz band in particular has been the focus of many academic research projects, industry prototyping and field trials. It has the capacity to accommodate a range of high data rate applications, and there are no federal allocations in the band. It’s one of the bands queued to go up for auction on Nov. 14, to be followed by the 24 GHz band.
Verizon is by far the largest holder of mmWave licenses today, and according to T-Mobile’s estimates, Verizon has 76% of the 28 GHz spectrum in the top 50 markets. T-Mobile holds 12% and 10% belongs to the “other” category, leaving just 2% available for auction in the top markets. So, it would seem, that auction won’t take too long.
Some cellular industry stakeholders would like to see all the mmWave bands auctioned together. T-Mobile has argued that including the 37 GHz, 39 GHz and 47 GHz bands along with the 24 GHz band would foster more competition. The reasoning is that if more spectrum gets teed up faster, then manufacturers will direct their attention to technology development in those bands.
To be clear, well-respected engineers still point out mmWave’s limitations. They say it’s not ideal for mobility, and because it’s never been deployed on a wide commercial scale in the real world, their hesitations could ring true.
The tides are changing
However, Verizon and AT&T both have tested mmWave extensively in field trials and report that they were pleasantly surprised to find their earlier preconceptions were overblown. Beamforming, massive MIMO and other techniques are being applied to steer signals around obstacles or use them to bounce off of them. Massive MIMO has been likened to having a bunch of flashlights targeting a group of users rather than a single floodlight.
Verizon revealed last fall that its engineers had observed that millimeter wave propagates a little better than expected in terms of line of sight and elevation. They assumed they could get to the 6th floor of an apartment or office building, but their tests showed they were able to get up as high as the 19th floor, and they saw better penetration through various wall materials and latency under 10 milliseconds.
More recently, AT&T said that after years of research, it’s confident it has all the answers it needs to deploy a mobile 5G network that works for people all over the country. Its engineers observed no impacts on 5G millimeter wave signal performance due to rain, snow or other weather events, for example, in Kalamazoo, Michigan, and observed a latency rate of 9-12 milliseconds in Waco, Texas.
Engineers are taking mobility very seriously. Last year, Qualcomm Technologies announced it had achieved a 5G data connection on a 5G modem chipset for mobile devices—with an emphasis on “mobile.” Similar to the days when cynics challenged whether Qualcomm’s CDMA invention would work, the company said it would prove that mobility can be accomplished at millimeter-wave frequencies, something a lot of people said wouldn’t happen soon, if at all.
There are plenty of examples of mmWave spectrum being used by a variety of players. Facebook already built a sizable 60 GHz network in San Jose, California, and the 60 GHz ecosystem is growing. Concerns about oxygen absorption are no longer looming.
In New York City, a much smaller company called Skywire uses what’s called lightly licensed 70-80 GHz spectrum for a fixed wireless service, reaching neighborhoods that don’t have adequate broadband from other sources. And it’s not just urban areas. In a paper titled “Millimeter Wave Wireless Communications: New Results for Rural Connectivity,” NYU Wireless researchers described in detail the measurements and models that validate rural millimeter wave path loss models.
To be sure, a lot of spectrum bands are in play at the FCC. Commissioner Jessica Rosenworcel has identified open dockets in the 3.5 GHz, 3.7-4.2 GHz, 6 GHz, 24 GHz, 28 GHz, 32 GHz, 37 GHz, 39 GHz, 42 GHz, 47 GHz, 50 GHz, 70 GHz and 80 GHz, among others.
Meanwhile, the commission is seeking to unleash new spectrum in frequencies above 95 GHz—“way, way up there” spectrum that some see as going overboard. But FCC Chairman Ajit Pai said the point is the U.S. must be open to new technologies that haven’t even been developed. "While we don’t know precisely how far the laws of physics will permit us to go, we do know there’s potential and interest. Engineers and entrepreneurs need to have the ability to push the envelope," he said.
Pai noted that the skeptics have been proven wrong before. So, who knows how far engineers will push the envelope in the next 10 years?