Network Neutrality

This seems to be a particularly busy summer for consultations! I've just submitted a JANET(UK) response to an Ofcom discussion paper on Traffic Management and "Net Neutrality". The quotes are Ofcom's and I'm reassured to see them because I've always suspected the phrase of being something of a banner that can be waved in support of a number of different viewpoints about openness on the Internet.

So why is this relevant to a network like JANET, whose research and education purpose already requires it to be as open and neutral as possible? The first thing to catch my eye was the other phrase in the title - "traffic management", which is something we do. To avoid mutual disruption, high-bandwidth and experimental uses of JANET are given their own parts of the network through the JANET lightpath and aurora services. Even on the general-purpose parts of JANET we do block traffic to particular addresses or ports from time to time, as permitted on a temporary basis by paragraph 10 of the JANET Security Policy, to give connected organisations time to deal with new security problems that may affect them and others. Connected sites and networks may also use Quality of Service and other prioritisation technologies to make sure their networks deliver the services they need - for example where voice or video traffic shares a network with bursty traffic with less demanding performance requirements, traffic management may be the only way to make voice and video usable. This seems to me necessary, and entirely legitimate, to make networks work. My Ofcom response, therefore, stresses that traffic management technologies are essential, and that any regulation that may be proposed in future in the interests of "Net Neutrality" must focus on undesirable uses of those technologies, not on the technologies themselves.

About half of the discussion paper deals with issues specific to consumer networks, so the JANET experience doesn't seem relevant. However for my own interest I did read those sections to try to work out what Ofcom thinks are the problems within their definition of "Network Neutrality". The paper doesn't explicitly make this distinction, but I think there are two different issues:

• unfair competition, most obviously if an ISP offers an application service that competes with a third party service then the ISP might make it more difficult for their customers to access the competitor in order to increase their own business. At some (not always obvious) point that slips from legitimate to unfair competition. Fortunately there is already an extensive range of competition law and cases to decide when competition is unfair and a well-known "toolkit" of transparency and regulation that can be used to keep competition fair;
• the harder issue, because it's a question of what society wants from the Internet, is ensuring that the Internet remains a space that supports innovation and not just by existing providers. Commentators often contrast the Internet, where (at least in theory) any two users can get together and run a new application or protocol, with the old telephone network where both law and technology meant that innovations were impossible without the approval of the network operator. There's little question that that openness was critical for the rapid development of our current networked society - the World Wide Web was created by a "mere user" - but it's becoming apparent that IP technology doesn't automatically guarantee the ability to innovate and that there may be economic incentives for network and service providers to try to keep innovation to themselves. The really interesting Network Neutrality question, it seems to me, is how much society needs and wants to preserve that ability to innovate and how that can be done.