Ready for customers 2019-2021. John Chapman made cable engineering history with the first demonstration of full duplex cable at ANGA COM in Cologne, pictured. The setup doesn't look like much until you realize this is the world's first prototype of a 10 gigabit down, 5 gigabit up, cable system. By using the same spectrum upstream and down, remarkable speeds are possible.

Chapman outlined the path to gigabit cable in 2005. Almost no one believed it possible except for a handful of innovative engineers. His ideas became central to DOCSIS 3.0 and 3.1. Comcast today is offering the gig downstream to millions of customers. If John says FDX gigabit upstream will work, I'm a true believer. (The schedule isn't guaranteed, of course.)

Full duplex communication has a long history, but not at the speeds of today's wireless and modems. Phil Levis & Sachin Katti of Stanford have spun off Kumu Networks, which has ambitious plans for wireless FDX and inspired renewed interest.

FDX for cable is a harder problem. Cable systems are "point to multipoint" - one central unit, the CMTS, attaches to many homes. The challenge with point to multipoint is that an echo canceller can only be used at the CMTS end.

Peak Internet growth may have been a couple of years ago. For more than a decade, Internet traffic went up ~40% every year. Cisco's VNI, the most accurate numbers available, sees growth this year down to 27% on landlines and falling to 15-20% many places over the next few years. (Chart below) (Thank you, Arielle)

The result: bandwidth cost per month per subscriber will continue flat to down. For large carriers, that's been about $1/month since ~2003. Moore's Law has been reducing equipment costs at a similar rate. This is confirmed by the global carrier spending on service provider routers, which continues about flat per customer.

Mobile growth is staying higher. 40-50% worldwide. Fortunately, mobile technology is moving even faster. With today's level of capex, LTE networks can increase capacity 10x to 15x.

Frontier lost 107K broadband subscribers in Q1. For years, I've been expecting the low capex at Century, Frontier and Windstream would hurt, as their offering falls further and further behind cable. In most of the United States, cable's basic offering is 50-70 megabits down, 5 up. The three mid-sized U.S. telcos offer 5-40 megabits down to most, often at the lower end. Especially at Frontier, marketing has been deceptive. They advertise "up to 100 megabits" when the strong majority of their homes can't get even half that.

The future is now, perhaps. Frontier is down 107K in Q1, while Windstream and Fairpoint also fell.

Use the plastic sheathing around the phone wires as a waveguide. With 1,000x more spectrum, you get 1,000 more speed. Glass fiber optic waveguides can carry 250 terabits. John Cioffi wants to apply similar techniques using the air gaps between the plastic surrounding a billion phone lines.

The proposed terabit DSL can use 300 GHz+ of spectrum, "submillimeter waves." Current DSLs use 100-500 MHz. Higher frequencies just wouldn't make it through a standard copper wire. The signals get weaker (attenuate) very quickly in copper. 500 MHz can only carry about 30 meters. Gigahertz, even less.

Cioffi proposes using the tiny air spaces between the plastic insulated wires as a "waveguide." The signal would travel over the air gaps, not the copper or plastic. Fiber optic, glass or plastic, "guides" the waves; why couldn't the plastic insulation do similar? (The waves are very, very small. They can easily fit in the gaps.) The signal is carried through the air between the wires, not on the copper wires.

Nobody's built a demonstration unit yet, but numerous very respected engineers are impressed by the concept. Using submillimeter frequencies, the telephone binders as waveguides, and massive calculations for "noise cancellation," speeds as high as a terabit should be possible.

Those specifications are at or slightly beyond the state of the art, so building the first unit is still to come. With support, Cioffi believes it can be done in just a few years and commercially soon after. That remains unproven but the industry is alive with enthusiasm.  Technical presentation .

Terabit DSL Proposed By Stanford Professor John Cioffi

One Terabit 100 meters, 100 Gigabits 300 meters, 10 Gigabits 500 meters 

On May 10, 2017, at the Paris G.fast Summit, John Cioffi will describe a path to a terabit over copper. That is not a typographic error, although John's ideas are far beyond the state of the art. G.fast using 106 MHz is now reaching a gigabit. 212 MHz is soon to ship, with correspondingly high data rates. Alcatel has pushed into higher spectrum to demonstrate 10 gigabits. Physicists and astronomers work with submillimeter waves from 300 GHz to 3 terahertz. Nothing like that is used in communications

Wireless systems use much higher frequencies. 28 GHz & 39 GHz 5G millimeter systems are just getting to the field, with millions likely to ship next decade. 70 GHz backhaul systems are in commercial production. Even faster wireless systems are in the labs. Submillimeter waves begin at 300 GHz and go up to 3 terahertz. While speeds like that have not been used in communications, they are common in physics and astronomy. 

I'm pretty sure it won't be that high, but with "unlimited" wireless it should become more of a factor. Parks Associates, a decent research house, just put out a survey, "10% of U.S. broadband households likely to cancel their fixed broadband Internet service over next 12 months. ... 15% of heads-of-household ages 25-34 are likely to cancel their fixed broadband service in the next 12 months."

Wireless broadband is now fast enough many places for almost all practical purposes. Much of the U.S. is covered at 20 meg or more, higher than most DSL. That's enough for 5 HD streams. The speeds in many locations are headed to 50-150 megabits over the next two years. 

The limiting factor has been the cap, often at 2-10 gigabytes.

Estimate for 100% ultra-high speed goes from € 20 billion to € 34.9 billion. Cour des Comptes, the French government auditor, has recommended a rethink of France's plan to bring very high speed to 80% and high speed to 100%. They estimate the cost will be much higher and the expected private investors are not interested. Instead, they call for "un « mix technologique »" 

Sebastian Soriano, chief of regulator ARCEP, warns of "rural desertification and the risk of decommissioning of certain territories, their inhabitants and their businesses." He recommends only modest changes to the plan, such as wholesale operators in rural areas. 

Time Warner gamed the FCC AskSam testing. The multibillion dollar New York State suit against Charter/Time Warner revealed details about how they delivered false results to the FCC tests, in a manner reminiscent of the $20B Mercedes-Benz emissions fraud. Hundreds of thousands of customers were overcharged $20, $30, and $40/per month, often for several years. Because TWC blocked interconnections with companies like Netflix, millions of customers did not get the Internet connection TWC promised and advertised.A 20% refund to those customers adds up to $billions. 

That I got this wrong is not nearly as interesting as the story itself, but I'm putting my correction first. My frequent comments about U.S. cable systems generally delivering promised speeds or close. I hope the other cablecos, especially Comcast, were more honorable.

Time Warner sold 50, 200, and 300 megabit service to hundreds of thousands of customers they knew had D2 modems that couldn't deliver the speeds. That's just simple fraud and they collected $hundreds of millions. A team of lawyers worked through an massive pile of subpoenaed documents and discovered smoking gun after smoking gun.

The Attorney-General's suit used Time Warner's own data to show customers often got 30%, 75% and even 90% less than the speeds they were charged for. They also found definitive proof that TWC actively refused to accept data like Netflix streams in order to force Netflix, Cogent, and others to pay Time Warner or be effectively cut off from millions of customers.

Why bother to offload when your LTE is fast and unlimited? Growth rates seven years ago approached 100% as people first got iPhones. They are much lower now. Verizon last year saw about a 45% growth in data per customer, confirming a trend to lower growth in traffic. Cisco (and I) have been predicting this fall for the last three years because smartphones are now nearly ubiquitous. AT&T and others have been telling Wall Street that customers with smartphones were only increasing usage by about 40%. (Trends worldwide are similar but details deserve another article. Some charts below and at Cisco)

Technology in wireless is moving so rapidly that telcos could increase capacity 8-10X in the next few years without raising capex. They may not because they won't be able to sell so much data. Cisco projects a 7X increase in demand in five years, less than the potential capacity increase. 

200M wireless, 2M wired. Reliance just signed up 100M 4G subscribers in six months; India is likely to pass the U.S. in broadband connections by the end of the year. About 15M of those have DSL, mostly from government owned BSNL & MTNL. Bharti, which is the fourth largest telco in the world, has just connected DSL customer two million. The current offering is VDSL, which they sell as the misleading, "up to 100 megabits." I haven't found any evidence of vectoring nor a breakdown of actual speeds or distances. Based on the British experience, that means most customers will get 10 megabits to 50 megabits. They call it V-Fiber, a marketer's fantasy.

Reliance Jio is currently running fiber home, designed for a gig but apparently mostly sold at 100 megabits.

Strongman rulers delivering better results. We like to think democracy and openness are the keys to expanding broadband, and some very prominent ideologues have been making that claim. The data show otherwise. Egypt, the current growth leader, is a vicious military dictatorship.  Vietnam remains a single-party socialist republic. The Chinese government has lifted hundreds of millions out of poverty, one of history's greatest achievements. The horrors of 1958-1975 are now in the past, but it remains an authoritarian state. 

Thailand, #4 among nations with 1M or more connections, remains under a military government. #5 Belarus makes the Soviet Union look like a model Democracy. Among smaller countries, the leaders are Sri Lanka, Kosovo, and Syria.  The growth rates for Q3 2016 of these countries ranged from 3% to 8%.

Much as we'd like to believe it, the evidence is clear broadband growth does not require openness, freedom of speech, public private partnerships, light touch regulation, or the other recommendations of the Washington Consensus. Nor is "multi-stakeholder" control required. These may be good things - I support most of them - but making stupid claims in public is not the way to achieve them.  

Similar to Germany, faster than France. A few years ago, I was amazed to discover that Russia had some of the fastest Internet in Europe. Akamai's State of the Internet Report is based on a massive data set and showed excellent speeds in Russia. Most are served from fiber to the basement, inherently faster than most DSL, the main European connection. Russians generally live in apartment buildings and FTTB was the natural way to build the networks.

According to Akamai's report, the Q3 average connection speed in Russia was 11.6 Mbps, about midway between England's 9.7 Mbps and Germany's 13.7 Mbps. Russia is ahead of Germany in peak speed measures. (Unfortunately, the Akamai report has important methodological problems. The relative speeds do correspond to other data I've seen so I think they are in the right proportions.)

The market is intensely competitive despite only having three players in many areas. Prices were always low and fell further in dollar terms as the oil price dragged down the rural.

Shows the danger of underinvesting when competitors are upgrading. Unicom is the small carrier in China, with only 264M mobile and 75M broadband subscribers. In 2016, they added 11M mobile customers but only about 2M broadband. China Mobile's 1.3M LTE cell sites and massive fiber home build is pulling ahead. In 2016, they added 27 million mobile subs (to 849M) as well as 22M broadband (to 77M.) China Telecom has been adding LTE towers at a rapid pace. In 2016, China Telecom added 17M mobile (to 215M) and 10M broadband (to 123M.) 

China Unicom did not match the investments of the others. They had no choice but to cut prices & raise marketing expenses. Apparently, the Chinese market knew this profit warning was coming. The stock was not forced down. (This week, BT was down $8B and Verizon down $10B.) I wasn't particularly surprised either. I have been watching the nearly incredible build by China Mobile. China Telekom responded but CU held back. It looked like they were burying their heads in the sand.

I do not pick stocks, especially from 12,000 miles away and without a good knowledge of the market.

So cheap, all of cable will go DOCSIS 3.1. DOCSIS 3.1 is designed to do a gig up and down, but only the downstream is widely deployed. Gig downstream cable is deploying rapidly, with Comcast and Cox expecting to cover 50% of the U.S. in the next 24 months. Until now, cable upstream doesn't match fiber or G.fast, running 20-35 meg at best. Upstream DOCSIS 3.1 is not yet deployed anywhere, although Comcast has announced they will begin in 2017. Others will wait for full duplex. Vodafone in Spain now has promised a gig across their entire network by the end of 2018, including upstream.

Until now, cable upstream doesn't match fiber or G.fast, running 20-35 meg at best. Upstream DOCSIS 3.1 is not yet deployed anywhere, although Comcast has announced they will begin in 2017. Others will wait for full duplex. Vodafone in Spain now has promised a gig across their entire network by the end of 2018, including upstream. I wouldn't be surprised if the pr is running ahead of the likely deployment and that the upstream is a little later.

Vodafone Spain is built around the former ONO, purchased by Voda in 2014 for €7.2 billion. It has 2.1M cable customers while passing 9.5M.

First major cableco to make the switch. The cost will be high but investors seem happy. The stock went up in a down market. I found the move particularly surprising because Altice is very heavily leveraged after paying well above market for Cablevision and others. Patrick Drahi is of the John Malone school: high leverage and risk, GAAP losses, prayer. That is working well in today's low interest environment.

Other cablecos trust DOCSIS 3.1 to remain competitive with fiber. Comcast is offering a gigabit downstream of DOCSIS 3.1 to millions of subscribers, with hundreds of megabits of DOCSIS upstream beginning in 2017. Cablevision faces Verizon FiOS, still America's best network although getting a little long in the tooth. I suspect Altice expects savings from a new, all IP network and ultimately lower operating costs. Fiber will enhance their strong share of the many businesses in Long Island. Once the decision was made to go all fiber, offering 10 gig was logical. The difference in equipment cost is small. 

Liberty Global is using fiber home for as much as 2/3rds of the new homes they are passing and many other cablecos are installing more fiber. As far as I know, Altice is the first large cableco to rip out a DOCSIS network and bring in fiber home. (Calix is providing gear for some small cablecos.) 

Altice wants to IPO the U.S. division, cash out a portion, and use stock to acquire more companies.

Moving from G.hn to G.hn Wave 2 at "up to 1 gig." G.hn has grown beyond home powerline into a legitimate contender for in-building broadband. KT, using Marvell chips has been offering hundreds of megabits since late 2014. KT reaches several million apartments now and intends to cover 95% of Koreans who can't get fiberhome by the end of 2017. That puts G.hn a year or more ahead of G.fast.

KT, working with Ubiquoss and Lightworks, has begun exporting the system. They have a deal with Turk Telecom and a  demonstration building in Barcelona near Mobile World Congress. As you can see from their pr below, they plan a strong export campaign; the picture is the cover of their slick brochure.

G.hn is not as complex as G.fast with vectoring and many other features. That means it's simpler and cheaper. Chano Gomez of Marvell believes Wave 2 can match today's G.fast chips. I haven't seen test data from an independent source.

Jackie Joyner-Kersee, Comcast's fastestJorge Salinger is ready as soon as the vendors can deliver.  Casa Systems has already demonstrated 400 megabits upstream and will soon show 600 megabits. They will only offer gigabits downstream in 2016, Daniel Frankel reports, “because 3.1 was not available in upstream. ... I think it’ll be in 2017.” Comcast's $70 gigabit downstream is available today in parts of Chicago, Nashville, and Nashville with a promise to cover over 40M homes by 2018. The upstream is limited to 35 megabits, but that's about to change. 

I had heard from cable people that they would wait for Full Duplex before upgrading speeds, so this is good news. Full Duplex - sending both upstream and down in the same spectrum - is an active project at CableLabs but likely 3-7 years from volume deployment. Comcast instead will use dedicated spectrum for upstream, possibly eight 6 MHz channels. AT&T is serious about adding 12M "gigabit" homes, using GPON and G.fast, and Comcast wants to be able to match the gigabits soon available to about 30% of the U.S.

Cable systems are about to get a massive capacity improvement with software defined CCAPs running on easily upgradable hardware. Harmonic has just entered the market with a software-defined unit running on a common, off-the-shelf Intel PC.

In large volume, 2/10ths to 5/10ths a cent. That suggests large ISPs pay something less than a penny per gigabyte. If you use 139 gigabytes/month, that costs your provider something like $1/month. (Doubling transit costs gives you a rough estimate of the cost to the carrier, which also has to carry the bits to your local exchange.)

Akamai recently bid "$0.002 per GB delivered," according to Dan Rayburn, a streaming media expert I trust. Dan notes that's the lowest price he's ever seen for a content delivery deal, and other recent bids have been as high as half a cent. Prices have been driven down by large customers, including Apple, doing their own content carriage. Amazon is in the game today as well.

Hurricane Electric is advertising Internet transit at dozens of locations for $0.20/Mbps, which works out to a similar price.

40% last two years, 30% expected next two years. NBN Chief Bill Morrow sees the Netflix surge passing and future growth slowing. The average NBN user draws 131 gigabytes/month, reports Geoff Long at CommsDay. (Not linkable. CommsDay is an excellent news source, available only via a professionally priced subscription.) 

Until two or three years ago, landline data traffic for the average user had been growing at 40% +-5% for a dozen years. Fortunately, Moore's Law has brought down the costs of the necessary components at a similar rate of about 40%. The marginal cost per subscriber per month has stayed flat at all the large carriers: About $1/subscriber/month across the developed world. Wireless growth has been similarly slowing.

Shared networks can work remarkably well. Across the U.S., AT&T, Comcast, and Verizon are in a fierce marketing battle to offer fiber to new developments. Table stakes now are a reliable gigabit network and premium TV offerings. 400,000 new apartments go up every year. They add less than 1/2 of 1% to the housing stock but are a prime growth market. 

I suspect there's a slight exaggeration here, but 97-99% would make most of us very happy. U.S. & U.K. government testing (SamKnows) have long demonstrated that most cable networks are darn close to 100% delivery. In 2014, FCC tests showed 95% of Comcast customers received between 109% and 119% of advertised upload speeds.

If cable/DSL is good, people won't switch. Kevin McLaughlin's article is unconfirmed but makes sense. Update Someone who knows thinks this is false. The analysis stands whether or not there are cuts today.End update.

 I have 200/20 cable; Jennie 50/50 fiber. Both are rock solid and uncongested. For most people that's fine and Google's gigabit isn't worth switching for. Reported result: Cost per new customer blows out the economics. Fiber needs to win ?30% to 50% of the market. That's hard if the existing carriers aren't so bad. LTE by the end of this year will be at a gigabit, (shared) with more antennas likely to allow servicing more fixed customers.

LTE by the end of this year will be at a gigabit, (shared.) Many antenna MIMO will allow servicing more fixed customers even before 5G is ready. Comcast and Cox are promising a gigabit (shared) to half the country within two years.

People hate to switch.

AT&T cutting 80K jobs. Carriers are gung-ho for SDN/NFV because it will be more efficient. Tens of thousands of people will lose their jobs at telcos, cablecos, and suppliers. In the long run, the economy benefits from the increased productivity, as do shareholders. But many of those fired will never get a decent position again, especially older workers. 

There are no easy answers. AT&T is paying for extensive employee retraining, hoping linemen can learn to be programmers for cloud computing. They donated $million to Georgia Tech's innovative fully accredited online Master's in Computer Science and are paying tuition for employees. Some will be able to make the change. 

Nokia Will Make Massive Layoffs and Ericsson plans to lay off thousands are recent headlines. None of these are related directly to SDN deployments, of which there are very few so far.