Millimeter Wave: What’s Goin’ On

Verizon has the only substantial mmWave network on the planet. They may have already covered 5% of the US, but most is not yet turned on. The company refuses to release any figures other than 30 million passed one day, about a quarter of the US. The majority of Verizon’s “5G” will be low or midband, which will cover about 40% of the US next summer.*

At Verizon, business areas and some neighborhoods will get mmWave, the good stuff designed for a gig or so. The rest will get low and mid-band, details pending CBRS and C-Band auctions. It will mostly be 70%-90% slower. Guess which neighborhoods will get the slow stuff.

KT was one of the early pioneers in mmWave, supported by Samsung research. Jerry Pi at Samsung was one of the first in 5G mmWave research and Samsung the first to put a large team of engineers to work building actual equipment. KT nows talks vaguely of sometime in the future, probably highly limited for the next several years.

NTT DOCOMO has similar plans. CTO Seizo Onoe has been involved in mmWave since the beginning. For now, DOCOMO is only talking about limited trials for several years. It may try for a splash at Olympic venues. Telefonica Deutsche had plans but seem to have set them aside.

Few are building mmWave. Most telcos believe they have more capacity than they can sell until the middle of the decade or later

Traffic growth is no longer exponential. In many of the developed countries, traffic growth is down to 25%-35%. There are a limited number of hours every week to watch Netflix.

The 100 MHz many telcos now have at 2.5-4.2 almost doubles capacity. The many antennas of Massive MIMO roughly triple it again. Verizon and AT&T in 2018 were using less than half their low-band spectrum, another reservoir of potential bits. SON, CA and other technologies are improving.

Between about 2016 & 2020, telco capacity has been able to grow 12 times or more without raising capex. The tech is that good. With traffic growth slowing, I calculated in 2018 that most telcos wouldn’t need mmWave until well into the next decade.

Verizon CEO Hans Vestberg believed that the extraordinarily low cost per bit of mmWave would be a huge advantage. I’d like to think he was right but the results aren’t there yet.

AT&T 5G will include a modest amount of mmWave. Stephens notes

we have 2 gig speeds.They will be focused on; first, business applications, automated factories, campus capabilities, entertainment capabilities in a sporting venue like an AT&T Stadium. We’ll focus there then to other applications. We will also use it for really high traffic areas. So in our campus in Downtown Dallas, we will have 5G capabilities and we’ll use it, try to offload a lot of the traffic that comes with so many people being in one location.

Korean and Japanese telcos have also said they will do a modest amount of mmWave in certain locations, but little is in the field. The Chinese see 2020 & 2021 as trial periods for mmWave.

* President Ronan speaking to investors. Verizon otherwise tries to hide these plans.