ITU's IMT-2020 defined 5G. 2030 could have the same effect on what comes next. The group welcomes new members. If you understand the issues and can present them articulately, you will be heard. October 2nd is an open workshop, the 3rd and 4th a Focus Group open to new members.

I've been surprised how much impact a thoughtful person can make at ITU, If you can't come to New York, written presentations are welcome. The number submitted hasn't been overwhelming so they will be read.

Below are some suggestions I made about keeping costs and royalties down, sharing spectrum including with Wi-Fi, and promoting competition. Historically, Latin America, Africa and South Asia have been under-represented. We must include their needs, Most of Africa is off the grid; a low energy option is required. 

Based on the results of auctions this year in Spain, England, and Korea, the 3.7-4.2 GHz band would go for $10B-$25B. This was little used until 64 antenna Massive MIMO has suddenly made it very valuable. Antennas in higher frequencies can be smaller, and tens of thousands of 64/128 antenna rigs have shipped. The first results are customer speeds in the hundreds of megabits, confirming the trials and models. There's enough spectrum involved that all the telcos could double their spectrum holding.

In Britain, 150 MHz of spectrum went for 1.05B pounds, about US$1.35B. At that rate, 500 MHz would be $3.3B. The U.S. has almost five times Britain's population, so ~US$16B. None of the four UK carriers got more than 50 MHz, less than ideal. Bidding was strong.

In Korea, 280 MHz went for $2.7B. Adjusting for the U.S. population, that would be ~US$30B.

In Spain, 320 MHz went for US$511M. Spain had only three bidders and enough spectrum for all to get 90-100 MHz, the ideal for the band. So there was no competitive bidding and they got it cheap. 

The FCC is considering skipping an auction for the 3.7-4.2 GHz bands, key for 5G. The market cap of Intelsat, a licensee, has gone up from $500M to $3B, just on the hope Pai and Reilly will let this go through.

Lee Hicks, Verizon Vice President - Network Planning, reiterated that Verizon has been reducing cost per bit by about 40% per year and expected to continue to do so. That corresponds to then-CEO Lowell McAdam's comment that mmWave would bring down the cost per bit by 9/10ths. Verizon's profit and loss confirms this general trend: costs and profits haven't changed much while traffic growth remains high. That would be impossible without a large increase in efficiency.

Traffic growth is falling towards 30%, according to the Cisco VNI. Unless capex falls, the glut of capacity without a market will get worse. 

Verizon expects to save half by replacing ~200,000 old routers, etc. with new, more efficient, standardized gear. They are active in SDN/NFV, including some pioneering work. CEO Hans Vestberg said that 5G equipment prices are coming down so fast it can afford to cover "the whole country" in much less than ten years. (Comments like that usually mean 85% to 95%.) He hated the price drops when he was at Ericsson and loves them now that he is at Verizon. All telcos are seeing dramatic falls in costs.

The other U.S. and Canadian telcos are also seeing a decline. The implications include:

Pal Zarandy, an always interesting analyst, tested Elisa's "world's first" 5G network from inside his office and outside with line of sight. The results were as expected: 5G at 3.5 GHz and Massive MIMO performs about the same as a good 4G network. "Gigabit LTE" - with tested speeds in the low or mid hundreds of megabits - has been shipping since 2016. Adding a minor software tweak, NR, and calling it "5G" did little to improve performance. (The battle to limit the term "5G" to true high-speed millimeter waves has been lost, unfortunately.) This is the dirty secret of "5G." Almost all the claimed new uses can be met with 4G. Anyone who doesn't know that should ask an engineer.

The Elisa results are important because they confirm that Massive MIMO makes 3.5 GHz spectrum practical to use. Pal writes:

"I think this is the key: 'Compared to gigabit LTE, the game changer features of the 3.5 GHz band won’t be single user speeds but instead the aggregate mobile data network capacity. This we couldn’t (and were not planning to) test, but Elisa’s network experts confirmed our views.'

Using 64 transmit antennas, Elisa in 3.5 GHz spectrum is getting throughput in a range similar to 8 antennas in 1.8 GHz. Without far more data, I can't be more specific than that. It may turn out that 100 Mhz in 3.5 GHz delivers about the same capacity as 60 MHz in lower bands. Because higher frequency antennas are smaller, even 128 antennas can be a reasonable size and not prohibitively expensive. 

Between 3.5 GHz and 4.2 GHz is enough spectrum to roughly double the capacity of today's networks.

Almost everybody thinks 5G mmWave will be very expensive, require vast numbers of small cells, and may not pay off.  That's all nonsense, according to Verizon CEO Hans Vestberg. He's done more testing of mmWave than anyone else except maybe Ted Rappaport and is probably correct. Here's what he said on CNBC:

5G is the same cost as 4G was. 

We designed 5G to make everything cordless.

Verizon is the first in the world. It's going to take some time to cover everyone but not a decade.

Spectrum is just one piece of the work. 

Stephane Richard at Orange/FT says the same, "I just want to remind everyone that our view is that 5G will not mean significant additional CapEx for us in the coming years. This is what we have constantly said and explained to the market, and this is still our view."

30M homes servable, 35% penetration in five years, significantly raising earnings. Brett Feldman made a gutsy call today. Brett has been accurate in the past about Verizon plans; he's the only other person who in Spring 2017 projected 30M. Brett's assumptions are. 

• Verizon will add 5G millimeter wave to all their existing cell sites. (I estimate 60,000-70, 000.) This brings about 30M homes out of incumbent territory within 2,000 feet. He expects that to be mostly completed by 2022.
• Verizon has demonstrated a gigabit at 2,000 feet, although I believe VZ  has provided no data on what percent of customers within 2,000 feet will actually be gigabit capable.
• Based on a large rollout in 2019 going forward and matching the 35% to 40% take rate for Fios, he calculates a likely 7.9M subscribers for 2023. Verizon today has only 7M broadband customers.
• Feldman expects Verizon mostly to upgrade existing cell sites rather than build many new ones in this period. He sees little increase in capex. I calculate that will allow Verizon to reach ~40% mmWave coverage, including incumbent territory. They will have "Gigabit LTE" elsewhere delivering speeds in the hundreds of meg. 
• $50/month is the most likely price but he expects Verizon to experiment. 

Add-on to come next year, at unknown price. It's the existing, not very impressive $480 Lenovo-Motorola Z3 and a promise of an add-on with a Qualcomm X50 modem. As you can see in the picture, the module is designed to snap on.

Ted Rappaport promises 5G phones by Xmas, but the quantity will be limited. The industry expectation has been "some" production units in the first half but quantities would be severely limited until the second half.

Verizon's suggestion these will be readily available early in the year is surprising. The whole 5G schedule has already been pulled up by a year, an extraordinary achievement in an almost unbelievably complicated product.

Modular, upgradeable phones like this have been around for a long time. They've never found much of a market. You generally take a performance hit if you only upgrade part of the phone. New phone prices kept coming down; if you needed a new function you probably bought a new phone.

A brilliant pr move calling this a 5G phone when it can't do 5G.

 "China is the exception; political forces are driving a huge nationwide deployment at incredible speed. We ran the numbers, and they will surprise most of the industry. We had been limiting our forecast based on capacity issues in the supply chain, but with big recent investments in component capacity, we have bumped up our forecast for 2019." Joe Madden, below. Madden's sources include major suppliers in his report, Semiconductors for Remote Radio Heads. Chip and component people need to know 6-12 months before the build in order to ramp up. Usually, they know what's coming.

In March, I wrote 2M 5G Cells At 3.5 GHz for China Telecom. Chengliang Zhang in March said CT would build mostly 3.5 GHz in 2020-2025. Madden's sources say that's been pulled to early 2019. That makes sense; Ericsson and Huawei are shipping production quality gear. Qualcomm's small RF front end looks like a breakthrough. While Ted Rappaport promises phones for Xmas, there may only be a few hundred of them. Word on the street is the phone ramp is late in the year. 

As far as I know, China is primarily 3.5 GHz. Six months ago, China Mobile told me mmWave would be limited until ~2022. Same for Japan. It's plausible they have changed the timeline. The gear is ready; Verizon probably has more than a thousand cells ready to turn on held up for marketing reasons.