1 Nuts and Bolts of Network Neutrality1 Edward W. Felten Center for Information Technology Policy Department of Computer Science, and Woodrow Wilson School of Public and International Affairs Princeton University Version of July 6, 2006 [email protected] Network neutrality is a vexing issue. Proponents of neutrality regulation argue that the free, innovative Internet of today is threatened and government action is needed to protect it. Opponents argue that regulation is not needed, or will be flawed in practice, or is a bad idea even in principle. One of the reasons the network neutrality debate is so murky is that relatively few people understand the mechanics of network discrimination. In reasoning about net neutrality it helps to understand the technical motivations for discrimination, the various kinds of discrimination and how they would actually be put into practice, and what countermeasures would then be available to users and regulators. These are what I want to explain in this essay. It’s not my goal to answer every question about net neutrality—that would require a book, not an essay. What I want to do is fill in some of the technical background in a way that illuminates the core issues, in the hope of providing a little clarity to the discussion. 1 Intelligence at the Edges vs. in the Middle The Internet consists of a set of end-user computers connected by infrastructure that carries data between those computers. This infrastructure is basically a set of routers (think: metal boxes with electronics inside) connected by links (think: long wires). Packets of data get passed from one router to another, via links. A packet is forwarded from router to router, until it arrives at its destination. 1 This paper is available online at http://itpolicy.princeton.edu/pub/neutrality.pdf. Thanks to Alex Halderman and Cameron Wilson for feedback on earlier versions.Nuts and Bolts of Network Neutrality Edward W. Felten 2 The Internet is unusual among networks in putting most of the intelligence in the computers at the edge of the network, rather than in the infrastructure at the heart of the network. The routers in the middle forward packets with only minor processing—all the heavy lifting takes place on the transmitting and receiving computers. This approach of putting intelligence at the edge of the network is known as the end-to-end principle, and it is one of the keys to the Internet’s success thus far. Putting the intelligence in the edge computers has several advantages. (1) Edge computers account for most of the devices involved in the network, so the edge computers collectively have most of the memory and processing power available to the network, and it makes sense to put the intelligence where these resources are available. (2) Edge computers have a better idea what the network’s users want, because they are owned and controlled directly by users. (3) Innovation usually happens faster at the edge of the network. In a sense, the net neutrality debate is a fight between the edges and the middle over control of the network. Neutrality regulation is generally supported by companies that provide services at the edge of the network, and is generally opposed by companies that manage the middle of the network. Each group wants the part of the network that it controls to have most of the intelligence, because more opportunities to innovate—and profit from innovation—are available to those who control the intelligent parts of the network. 2 Minimal vs. Non-minimal Discrimination Focus now on a single router (in the “middle” of the network). It has several incoming links on which packets arrive, and several outgoing links on which it can send packets. When a packet shows up on an incoming link, the router determines on which outgoing link the packet should be forwarded. If that outgoing link is available, the packet can be sent out on it immediately. But if the outgoing link is busy transmitting another packet, the newly arrived packet will have to wait—it will be “buffered” in the router’s memory, waiting its turn until the outgoing link is free. Buffering lets the router deal with temporary surges in traffic. But if packets keep showing up faster than they can be sent out on some outgoing link, the number of buffered packets will grow and grow, and eventually the router will run out of buffer memory. At that point, if one more packet shows up, the router has no choice but to discard a packet. It can discard the newly arriving packet, or it can make room for the Take-home lesson: THIS IS PARTLY A FIGHT TO CONTROL INNOVATION ON THE INTERNET.Nuts and Bolts of Network Neutrality Edward W. Felten 3 new packet by discarding an older packet waiting in the buffer, but something has to be discarded.2 When a router is forced to discard a packet, it can discard any packet it likes. One possibility is to assign priorities to the packets, and always discard the packet with lowest priority. 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. With minimal discrimination, if the network is not crowded, lots of low-priority packets can get through. Only when there is an unavoidable conflict with high-priority packets is a low-priority packet inconvenienced. In contrast, there is another, more drastic form of discrimination, in which routers discard some low-priority packets even when it is possible to forward or deliver every packet. A router 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. One of the basic questions to ask about any network discrimination regime is whether it is minimal or non-minimal in this sense, and one of the basic questions to ask about any rule limiting discrimination is how it applies to minimal versus non-minimal discrimination. We can imagine a policy, for example, that allows minimal discrimination but limits or bans non-minimal discrimination. This distinction matters, I think, because minimal and non-minimal discrimination are