September 21, 2007 8:31 AM
Maybe Internet Isn't Growing As Fast As Widely Thought...
Much has been made about the oncoming exaflood of Internet traffic, which threatens to overwhelm broadband networks, but at least one pundit is wondering what all the fuss is about.
Andrew Odlyzko, director of the University of Minnesota's Digital Technology Center, has been studying data from 100 public data traffic hubs around the world. His research shows that while Internet traffic may have been doubling in the early years, recently it has only been growing about 50% per year.
In this Startribune article he suggests that the warning cries of an overwhelmed web are overblown. (If you want to get into the specific numbers, check out his results here.)
This belief leads him to suggest that efforts by network operators to explore new business models for the delivery of Internet traffic to not be in preparation for overwhelming demand but instead merely an effort to exert more control over the Internet.
While his research is both interesting and valuable, I have a few issues with his conclusions.
First off, Odlyzko's claim that the lack of overwhelming demand equates to there being no need for managing bandwidth allocation--essentially an argument in favor of net neutrality--highlights a pervasive misunderstanding of how this issue, and government regulation in general, tends to work: passing net neutrality means infringing on the freedom of network operators to run their businesses, just as not passing it holds the threat of doing the same to Internet entrepreneurs.
As with most polarized debates around government policy, it's my belief that the ultimate answer isn't picking between an either/or situation but finding the best solution for encouraging the freedom and growth of both sides.
Secondly, while I don't dispute his numbers, I do take issue with his claim that since the Internet hasn't been doubling that there no longer is any problem with the exaflood. In fact, he even contradicts this thought in the same article: "If everybody rushed to download video from the Internet, it couldn't be done," he said.
I had thought we were all shooting for the same goal of eventually all forms of data transmission going over the Internet. If that's the case, then how can we not have a capacity issue if even today we can't support the delivery of video-on-demand to everyone, let alone all the other broadband applications that demand bandwidth, and not to mention the millions of people who have not yet discovered the Internet but hopefully will in the near future.
All this being said, I do heartily agree with two other assertions he makes.
For one, that quote I cited continues on to include his belief that people's habits aren't easily changed, and that's a primary limiting factor in the increasing demand for bandwidth. This is absolutely true. People often don't warm up to change right away, and for the most part the Internet is a highly amorphous and therefore uncertain and somewhat scary place. Even my beautiful wife is still wary of doing something as low level and proven as buying a book from Amazon.
Yet at the same time there's nothing really stopping everyone from trying to log onto the Internet simultaneously, and all it takes to initiate a mass migration like that would be another singular event, like another Katrina, or something more entertainment-related like letting people watch the Super Bowl online.
All of these possibilities are both likely to happen and certain to drive massive amounts of demand for bandwidth.
The other major point with which I couldn't agree more strongly is Odlyzko's assertion that our focus should not be on restricting Internet usage but instead on encouraging greater use of it.
That's precisely what I've been arguing for in this blog for the last six months.
The Internet has accomplished many great things, but there's still so much potential left untapped.
We need more people online. We need the people who are online to use it more. We need to take bold steps forward towards realizing the full potential of the Internet rather than standing around wringing our hands about demand for bandwidth overwhelming supply and making grandiose speeches about how impactful the Internet has already been when we've really only scratched the surface.
All this being said, it's my belief that realizing a fully Internet-enabled future demands we focus our attention equally on increasing both supply and demand for bandwidth. There's little doubt we need more bandwidth, created largely by the deployment of fiber. And we must be making every effort to drive up demand, through education, awareness, and supporting those companies that help make the Internet an interesting place to be.
My reason for pushing harder on the demand side than supply is that there's been a lot of talk around the need for more broadband over the last few years, but not enough about the need to continue encouraging demand. So it's rather heartening to see the debate beginning to turn that direction.