Suddenly 2017 rather than 2020 looks possible for Massive MIMO and beamforming. Verizon told Wall Street they were close to trials. Nokia next quarter will test Antonio Forenza's pCells, with similar features. Verizon's Lowell McAdam also briefed the street on a plan for "Staying ~two years ahead of competitors in network performance." (Paul de Sa.) They can't reach that goal without pulling up technologies that most thought were five years away.  

"Go Massive" is the conclusion of Robert Heath, at The University of Texas at Austin.

Hoping for 2017 although almost no one expected this before 2020 and "5G." CEO Lowell McAdams told analysts Verizon plans  "Staying ~two years ahead of competitors in network performance."  Massive MIMO and beamforming deliver "Huge gains in spectral efficiency." Delivered wireless can increase two to ten times over the next few years using arrays of thirty-five to hundreds of antennas. A 10x jump wouldn't surprise any of the researchers. Pulling that far ahead is probably an impossible dream. AT&T has almost caught up with Verizon. 

Until this week, almost everyone thought substantial Massive MIMO deployment unlikely until 2020 and 5G. The excellent Bernstein analyst Paul de Sa was the first to report "VZ will soon begin market trials of 5G capabilities including 'massive MIMO' (which is when a large number of antennas are packed into a single device) and beam forming (which is when a wireless signal is concentrated on a specific location)." I confirmed with a second analyst what was said.

Beginning trials soon would suggest deployment in 2017 if they go well.

10x or 100x more wireless throughput using many antennas. Masayoshi Son at Softbank has just deployed the first commercial Massive MIMO systems. 100 systems, each with 128 antennas, are being deployed in Japan. The early results show 6-10X capacity improvements. Some of the world's best engineers - Henry Samueli, Andrea Goldsmith, Arogyaswami Paulraj, Vint Cerf - expect ultimately MIMO will yield a 50X improvement. (That's not guaranteed, of course.)

Ericsson announced they'd be shipping 64 antenna systems in 2017. Huawei, ZTE, and Facebook have demonstrated 96-128 antenna systems. Each transmitting antenna sends a separate signal. The receivers mostly have two or four antennas. Each signal bounces off walls and other obstacles in a slightly different way, allowing the receivers to separate them. 

Paulraj invented MIMO in 1993 and nearly twenty years ago predicted a 100X improvement in capacity at very little power increase. Many of us have seen a preview of what's possible from WiFi 802.11ac. That uses 4 antennas and triples performance under the right conditions. Add more antennas and enough processing power and almost anything may be achievable. 

Verizon and Nokia are ready to test. Nokia CTO Hossein Moiin intends to test Antonio Forenza's pCell technology early next year. Last year, I wrote extremely skeptically about Artemis/pCell informed by three very respected engineers. CEO Steve Perlman made wildly implausible claims, including that they would deploy across 350 San Francisco rooftops by the end of last year. They haven't been seen.

Moiin is a respected, independent engineer. I have to look again. USC Prof Giuseppe Caire now says "The early trials showed pCell achieving far higher concurrent user capacity than any wireless technology I am aware of." That's very different from his earlier comments (below) and I assume he's been shown something newer. Peter White at Faultline has an important article about semi-secret trials in U.S. sports stadiums. (Paywall.) 

Verizon in a startling move Monday told analysts they would soon test Massive MIMO and beamforming. (pCell is very similar.) These are key "5G" tools that weren't expected until about 2020. 

Essentially, beamforming focuses the radio signal. It's concentrated and stronger to an individual receiver such as a mobile phone or WiFi. It's sometimes described as "steering" the signal. It's been used in radar and sonar for decades. More recently, it's common in 802.11ac WiFi and built into some wireless standards. As processing power becomes cheaper with Moore's Law, more powerful systems become practical. 

Eric Geir offers a lay explanation: "Wireless routers (or access points) and wireless adapters that don’t support beamforming broadcast data pretty much equally in all directions.

AT&T, Verizon cram 3-5 as many people on each cell site. AT&T has 50,000 cells, China Mobile 1,000,000, SK Korea 35,000, Spain 33,000. Currently, Albania, The Maldive Islands, and at least a dozen other countries have faster LTE networks than the USA. The chart is adjusted for population and AT&T obviously stands out. Verizon would be similar.

Since 2000, U.S. wireless networks have been far less reliable than many in Europe and East Asia, except for a brief period around 2010-2011.

NYU Wireless has just reported on 14,000 tests using 28 GHz and 73 GHz indoors. The paper isn't online yet so I put the abstract below. Professor Ted Rappaport believes, "These high frequencies will be an effective substitute when today's Wi-Fi frequencies get crowded."

The FCC is about to set aside some high frequencies for telco use. My opinion, not Ted's is that monopoly spectrum is obsolete. Wi-Fi is proving sharing is possible and productive.  Some monopoly spectrum is needed where reliability is important, but the 100 MHz Sprint, AT&T and Verizon each have is plenty.

Ted makes an important point.  "I think the FCC would do well to also authorize unlicensed bands in the near vicinity of that new mmWave spectrum.

Most experts thought it would require fiber. Karolina Wikander is confident. "Microwave backhaul technology is already able to handle 100 percent of all radio access sites' capacity needs, both now and in the future. By 2020, the technology will have evolved to support multi-Gb capacities in traditional frequency bands and beyond 10 Gb in the millimeter wave."

Most plans for 5G assume extraordinary reliability. The telcos want to control traffic and connected cars. Failures there - including in massive snow or rainstorms - can cost lives. Controlling the cars may require 1 millisecond latency. One expert thought that would be impractical without fiber. (Dan Warren, a good engineer, wonders whether 1 millisecond is practical no matter the technology.) Update - Ericsson pointed out that microwave latency is low. See below.

Millimeter wave high frequency for the 5G network itself is deeply troubling for policy.