|CWA, FTTH Council Call For Massive Cable Modem Buildout (Satire)|
|Friday, 19 December 2008 18:35|
In a surprise move September 23, the Fiber to the Home Council, the Communications Workers of America, Senator Jay Rockefeller and Congressman Anna Eshoo called for a massive upgrade of U.S. cable modem service, which competes directly with fiber to the home buildouts and the telcos that employ most CWA members. They supported massive subsidies to 100 megabit service. DOCSIS 3.0 is capable of offering 100 megabit service at about a tenth the cost of building fiber and is already deployed to over 15 million homes in the U.S. today and 20 million more in other countries. It's the natural way to get to 100 megabits if that the only criteria. Fiber is more reliable and expandable, but those are not included in the resolution.
"Rockefeller-Eshoo resolution (S. Res. 191 and H. Res. 1292) would call for the following:
As you know, a critical first step in developing America’s broadband strategy is to envision where we want to go. By establishing these ambitious national goals, Congress can communicate to the nation a robust vision of telecommunications infrastructure which will encourage a vibrant economy and enhance the social welfare of all Americans.
Setting actual speeds is a good way to define a broadband goal, so the form of their resolution is right on target. (Personally, I believe fiber is the better long term way to go when practical, because it is more reliable and upgradeable in the long term. I wouldn't spend tens of billions more of government money to add fiber where cable is already likely to offer 50 megabits without any subsidy whatsoever. That's over 75% of the U.S. in a few years. db)
Approximately 90% of U.S. homes can get 10 megabit per second download today via their cable modems, so the first goal is very easy to meet except for the last 1-3% of U.S. homes. Over 200M homes around the world can already get that speed from cable modems, probably more than than receive 10 megabits from DSL today.
For less than $100 per home, all those lines can be upgraded to DOCSIS 3.0 at 160 megabits-per-second downloads. That is more contended than most DSL, although the early results are promising. The 160 megabit chips are rapidly being superseded by 320 megabit chips, and 400 megabit units were demonstrated by Cisco/Linksys in 2006. We can't be sure until those units are widely deployed, but based on the experience to date with cable modems, it's almost certain the 320 megabit units will deliver 100 megabits or more 90+% of the time. Neil Berkett of Virgin Media in Britain has announced he will be offering 200 megabit service in a few years.
As John Chapman of Cisco discussed in 2003 at Fast Net Futures, DOCSIS 3.0 is designed for a full gigabit (contended.) Two of the world's leading cable engineers did a study about whether DOCSIS 3.0 could meet the Singapore requirements for gigabit service. Their conclusion was the technology was practical, but in 2010 would be too expensive. Moore's Law is rapidly bringing down the cost of the equipment however, and will almost certainly make it practical to offer hundreds of megabytes by 2015. In addition, switched digital is becoming nearly ubiquitous in U.S. cable over the next few years, which will make the needed channel bandwidth available.
Very few trade associations or unions would give such strong promotion to their direct competitors.
Obviously, CWA and the FTTH Council believed the 100 megabit would be interpreted by D.C. as fiber only. The FTTH people should know better, because they are constantly are telling telcos they need fiber to effectively compete with DOCSIS 3.0. The CWA folks, mostly economists and lawyers, may suffer from beltway blindness and not understand that cable can also get to 100 megabits. Three very smart, respected policy people I spoke with this week in D.C. didn't realize that.