Friday, April 22, 2005

Big Broadband: "Providing Ubiquitous Gigabit Networks"

Here's something that some folks should find interesting: it's not only Baller who believes that necessary big broadband is being stymied by squashing the municipalities. And it's not only pie-in-the-sky futurists that think that broadband needs will continue to expand rapidly. It isn't an invented cadre of "socialists and communists" who think that encouraging municipal entry is absolutely critical to America's future.

It is, among many others, the Institute of Electrical and Electronic Engineers, the IEEE, the dominant engineering association and the standards setting guys. (You've seen the tag "IEEE standards?" Those guys.) This is a group that nobody could rationally place in the ranks of those accused of irrationality. These guys are the opposite of ideologues. What they care about, by nature and by training, is getting the job done right. That pretty much defines an engineer's job. And what a new white paper published by the IEEE-USA makes perfectly clear is that the group is growing pretty impatient with both the unrealistic hype around alternatives to a full fiber-optic buildout in the U.S. and the growing trend toward blocking municipalities and other "user-owned" (as they so eloquently phrase it) entities from providing what private, for-profit carriers will not.

The IEEE on the bandwidth problem:
A new generation of broadband, or —gigabit networks,— can mean significant benefits to the United States, but our nation must act promptly to ensure that such an infrastructure is ubiquitous and available to all. If we do not act, the consequence will be to relegate the U.S. telecommunications infrastructure to an inferior competitive position, thus undermining the future of our country's economy.
That's pretty clear. And pretty clearly the same thing that Baller was saying over at TechSouth the day before yesterday. You'll note that these guys are as precise users of words as they are of any other tool. The service should be ubiquitous --everywhere-- and available to all. They are talking about what policy types (and LUS) would call universal service.

Why the solution should be fiber:
Residences of the future are likely to expect 100 Mb/s to 1 Gb/s. Although such data rates may not be used continuously, that capability is essential for particular applications, as described in the text. Networks such as these are gigabit networks.

In stark contrast, current broadband based on copper wire or coaxial cable links, such as Digital Subscriber Line (DSL) or cable modem, has a nominal (and asymmetric) speed of, say, 2 Mb/s. Thus, typical data rates on gigabit networks range from 50 to 5,000 times as speedy as current broadband networks, which include Wi-Fi and third-generation mobile networks. ...
All these facilities are lower in cost, but also lower in capability, than optical fiber. Although fast technological progress is being made across the board, the copper-wire based alternatives cannot reach fiber speeds.
Cannot. Remember, these are engineers speaking to engineers. What they mean is that the physical constraints of copper are simply so much greater than that of fiber that a word like "cannot" is warranted. Copper will always be slower. As any engineer will tell you.

Why don't we have the necessary fiber for universal gigabyte networks? The IEEE will tell you plainly:
Extending optical fiber access to end users is progressing. However, it is slowed in part by the high cost of capital expenditures and in part by non-market and anti-competitive business actions (and inactions) by incumbent service providers.
Any confusion? No? Good.

Can we count on the incumbents? No. After a review of the incumbent's projects that note serious problems with less than gigabit bandwidth and a committment to asymetiric bandwidth that leaves end users in the role of passive recipients rather than equal participants in the network, we read that even these inadequate initiatives:
...will be deployed where profitable, meaning “fiber to the dense” or, realistically, “fiber to the rich.” Again, doubtful profitability would foreclose penetration to non-affluent and dispersed U.S. premises throughout the country. Further, “the money” is in content, not carriage (except under monopoly conditions). So these initiatives rely for profitability on control of content by the network provider, rather than open access by competing service providers. Diversity of information would be limited. The result would be closed networks and restricted content, aggravating the digital divide and limiting the engine of innovation that could otherwise exist. In a word, these initiatives will fall short of providing an adequate nationwide gigabit infrastructure.
How to fix this situation? Well among other sensible ideas:
Reduce barriers to competition and deployment of user-owned networks, to facilitate continuing market restructuring in the public interest.
By "user owned" the IEEE means municipals, co-ops, and businesses and business consortia. Let the people who use it control their own destiny.

I've always liked engineers.

This white paper is really one you should read: Providing Ubiquitous Gigabit Networks.

(Thanks to Dirk van der Woude of Citynet in Amsterdam (www.citynet.nl) for the lead. Go take a gander over there. They're doing great things in Holland.)

No comments: