Thursday, September 4, 2008

Refocusing Internet Research?

I believe that the motivation for our reading Interdomain Internet Routing (Lecture 4) was to gain insight into the functional aspects of the BGP protocol. A little bit of searching shows that most of what has been published on BGP is of a technical nature.

As the beginning of this reading notes, however, the structure of the Internet--and in turn the BPG--is a function of the business model on which it is built. While everyone who writes about BGP is quick to point this fact out, they then proceed to tackle how the technologies facilitate the complex business relationships that exist among the various tiers of ISP's.

I wonder, is there a group of economists and business visionaries that are just as excited about the future of the internet, but perceive the problem from the inverse perspective? After all, technology isn't driving the internet, money is. The technology follows the economics. A lot has changed since the DARPA paper on TCP/IP was written. In those days, the objective of the Internet was to connect people that "needed" to be connected. At that time, researchers were inspired by their national spirit to build something in the name of National Defense. At least that is the romanticized version of the story that I have in my head. Is that how things really were?

Today, we don't even pretend like the Internet is a national icon, we openly admit that the Internet is a conglomeration of independent business entities (which represent their share-holders, not red-blooded American's everywhere) whose only motivation is to generate a profit.

I bring this up because, in light of this perceived transition (i.e. growing up?) of the Internet early in its formative years, I am surprised at the difficulty I have finding research into the economics that drive the development Internet. Rather than founding our guesses for the future of the Internet on things like Moore's Law, Bell's Law, or any other of today's technology trends, shouldn't we be listening more closely to Economists (and laws more like Metcalf's Law), whose job it is to study the nature and behavior of the System that sits at the core of the Internet, which is essentially a business system, not a computer system?

After reading the other paper that Randy assigned for today, I am even more convinced of the sentiments I expressed above.

What made this paper interesting was not the math or the techniques that they used, but rather the big picture question that they were asking, which nobody had even thought to ask up until that point. They asked: what are the relationships between AS's on the Internet? That this question remained unexplored until 2000 only supports my above observations that we were (and continue to be) focusing on the technology at the center of interest and the business model as the secondary interest, while it should be the other way around.

1 comment:

Randy H. Katz said...

Very insightful thoughts on your part -- the technical capabilities emerged first based on what the technologists thought might be useful -- and the business realities evolved later. Presumably, the next generation internet architecture must consider mutually suspicious entities that need to collaborate to deliver packets end-to-end. How exactly to do that is a research challenge.