March 08, 2006
Must-Read: Nuts and Bolts of Net Discrimination
Cable operators and phone companies have made great headway in warding off net neutrality regulations by reasonably arguing that it isn’t clear what net neutrality advocates want. Net neutrality advocates do indeed have a communications problem — proponents of mandatory net neutrality laws speak in generalities and have yet to advance a workable plan that makes sense.
Part of the problem for network neutrality advocates is the inherently technical nature of the Internet. Most of the proponents are political creatures with little technical training and it’s difficult for them to articulate persuasively just how broadband providers might mess with third-party services and applications.
However, that’s not a problem for Princeton computer science professor Ed Felten, who has produced a two-part simplified tutorial on the nuts and bolts of just how broadband providers might mickey with the network to make life miserable for third parties. In his first discussion, Professor Felten describes how the Internet is really a set of routers, which can easily get congested if too much traffic flows through them. When taxed to the point of buffer overflow, routers must drop packets.
So one form of discrimination that broadband providers can implement is assigning priority to the packets, a relatively benign problem that Felten calls minimal packet drop discrimination, which he concedes may be a fundamental engineering necessity.
When a router is forced to discard a packet, it can discard any packet it likes. One possibility is that it assigns priorities to the packets, and always discards the packet with lowest priority. The technology doesn’t constrain how packets are prioritized, as long as there is some quick way to find the lowest-priority packet when it becomes necessary to discard something.
This mechanism defines one type of network discrimination, which prioritizes packets and discards low-priority packets first, but only discards packets when that is absolutely necessary. I’ll call it minimal discrimination, because it only discriminates when it can’t serve everybody.
Less benign is what Felten calls non-minimal discrimination drops, where the broadband provider routinely assigns low-priorities to certain kinds of packets and discards them even when the router isn’t congested.
Contrast this with another, more drastic form of discrimination, which discards some low-priority packets even when it is possible to forward or deliver every packet. A network might, for example, limit low-priority packets to 20% of the network’s capacity, even if part of the other 80% is idle. I’ll call this non-minimal discrimination.
The second part of Professor Felten’s discussion deals with packet reordering, which allows a broadband provider to assign the order in which packets get routed. If the packets are in the router’s buffer, the router will decide which packets to send and when, and today this ordering is usually on a first-come first-served basis.
Just as is the case with packet drops, broadband providers can reorder the packets to create minimal delays and non-minimal delays, which inject what is called “jitter” into the transmission, a problem that most acutely affects voice and video transmissions.
A minimal form of delay discrimination only delays low-priority packets when it is necessary to delay some packet - for example when multiple packets are waiting for a link that can only transmit one packet at a time. There is also a non-minimal form of delay discrimination, which may delay a low-priority packet even when the link it needs is available.
Felten seems to argue that if net neutrality laws are adopted, the best method for ensuring neutrality would be to punish broadband providers that engage in the non-minimal forms of discrimination. But that, as he points out, is easier said than done.
A network provider who wants to cause high jitter can do so, and might have pretextual excuses for all of the steps it takes. Can regulators tell this kind of strategem apart from fair and justified engineering decisions that happen to cause a little temporary jitter?
Surely some discriminatory strategies are so obvious, and the offered engineering pretexts so weak, that we could block or punish them without worrying about being wrong. But there would be hard cases too. Net neutrality regulation, even if justified, will inevitably lead to some difficult line-drawing.
Posted by Cynthia Brumfield at March 8, 2006 09:47 AM
Comments
The issue being discussed, "router neutrality" if you will, isn't even an issue as far as I can tell. And, couldn't the affected content provider detect and deal with a concerted effort on the part of the pipeline owner? I would think the existing FCC mandates would have more than enough teeth to deal with the issue. Besides, if the transport providers face expanded regulation, we take away their incentive to innovate and to expand, and strengthen their networks. Congress should wait until there is an identifiable problem before getting involved.
Posted by: oldtimer38 at March 14, 2006 05:47 PM