Tuesday, Jan. 13, 2009 | The internet is the most complex and far-reaching system ever developed by man — without it, modern society would essentially grind to a halt. Yet it is about as regulated as a pick-up basketball game.
As a result there is no atlas, no global view of this vast system of interconnected computer networks that has become so crucial to modern daily life.
State and federal agencies keep track of how many cars travel on each segment of our road network. No one knows how many packets of information travel from node to node on the internet. The Federal Aviation Administration collects the name of every passenger who flies on a commercial airliner. No one has a sense of how much of each type of information — such as spam — that is being sent over the internet.
That so much about the internet is unknowable doesn’t necessarily mean that it is in trouble. But it does mean that if it were in trouble, we might not know about it until it was too late.
This is what really bothers Kimberly Claffy and Dmitri Krioukov, two internet researchers at the University of California, San Diego’s Supercomputer Center. Claffy and Krioukov are nationally renowned for their study of internet architecture, and their work has been widely published, most recently in November in the journal Nature Physics.
“We don’t have a rigorous discipline of internet science, no formal channel of getting data,” said Claffy, director of The Cooperative Association for Internet Data Analysis at UCSD. “Even economics has the Bureau of Labor Statistics — and [economics is] called the dismal science. So what does that make us?”
Frustrated, mainly. They say, as long as there is such scant oversight, the information highway is likely to get more expensive, less accessible and more vulnerable to viruses and cyber crime in general.
In their wildest dreams, the academics and U.S. Defense Department scientists who designed ARAPNET in the late 1960s did not envision YouTube, MySpace or iTunes. Yet their design remains the basic foundation of the internet.
And decisions were made throughout the internet’s development in the 1970s through the 1990s not to regulate it, and fierce debate has raged over those decisions for decades. The explosion of data — mainly video — traveling over the internet in recent years has brought these issues closer to a head.
This is why Claffy and Krioukov say there needs to be some kind of regulation. Claffy said a good start would be something akin to an international bureau of internet statistics, some kind of bird’s eye view of the entire system. Even the most basic data collection would allow researchers to see bottlenecks and other potential problems that can’t be seen now.
But in order for such an entity to be created, those who own the largest networks in the internet — internet providers like AT&T and Verizon — would have to agree to share their data, something they have no incentive or requirement to do.
“The global structure of internet is the cumulative result of local decisions by individual organizations (owners),” Krioukov said. “There are about 20,000 of these owners — and no one is requiring that these organizations play by a set of rules.”
Many researchers are concerned that the lack of regulation and the self interest of the large owners will combine to intrude on what is essentially the freedom of the internet.
For example, the overload of data the internet is experiencing requires bigger, faster and much more expensive routers. And it is likely that the large providers will pass these costs down to consumers. The result could be a tiered service model in which the level of someone’s internet access to depends on how much they are able to pay.
Others, especially business interests, say such advocacy for more regulation is a solution in search of a problem. Claffy believes the truth is somewhere in between. And she has spent her career trying to get at that truth.
The 39-year-old earned a bachelor’s from Stanford University in 1989, and received a Ph.D. in computer science and engineering from UCSD in 1994. Three years later she founded the data analysis association, dubbed CAIDA. Krioukov, also 39, is a native of St. Petersburg, Russia. He received a Ph.D. in physics from Old Dominion University in 1998, and came to work at CAIDA in 2004.
An hour long conversation with the two can cover just about all things great and small on the internet. Recently they’ve been working on great rather than small. The paper they co-authored for Nature Physics (along with Marian Boguñà, a researcher at the University of Barcelona, Spain) focused on the internet and how it relates to other complex networks.
The three researchers developed a mathematical model, called “hidden metric space.” The model offers an explanation on how both man-made networks like the internet, and natural networks, like the human brain, are similarly affected by what is known as the “small-world phenomenon,” in which two locations in any given network are, on average, no more than six steps from each other.
In popular culture it is illustrated by the “six degrees from Kevin Bacon” game — any actor can be connected to Kevin Bacon by six or fewer films. In the world of internet routing, it relates to the efficiency in which information packets go from node to node. Nodes are devices that can send and receive information.
The research suggests that the internet has an underlying structure similar to complex natural networks — such as a network of Hollywood actors, directors and producers.
If developed further, this research could be an answer for the rising number of so-called internet “black holes,” which happen when routers become so overwhelmed by information that they can’t keep up, and links to parts of the internet end up severed. These black holes are slowing the work of researchers and other high-level users of the internet. And researchers say if not dealt with, black holes will ultimately affect the average user.
When will this day come, and how bad will it be? Good question.
“The bottom line is the internet, like your brain, is a complex system. And the lack of predictive power, or knowledge, of where this complex system will be in the next year or less, is the fundamental problem of contemporary science,” Krioukov said. “You ask me what could happen to the internet — there are a range of possibilities.”