BEREC on Traffic Management

A short but interesting note from BEREC (the group of European Telecoms Regulators) on information they have collected from 400 mobile and fixed line operators on their traffic management practices. I get the impression BEREC were surprised at how varied the reports were! They only seem to have identified two patterns - “The most frequently reported traffic management practices are the blocking and/or throttling of peer-to-peer (P2P) traffic, on both fixed and mobile networks, and the blocking of Voice over IP (VoIP) traffic (mostly on mobile networks)” - but apart from that found “a very wide range of practices across Europe, and an equally wide range of implementation methods and policy justifications for them”. BEREC plan further analysis of the results, with two aims. First is to try to categorise types of traffic management so they can detect whether informing customers about it (Ofcom’s currently favoured option) results in any changes or trends that may need to be addressed by guidance or legislation.

The second, perhaps more interesting, is to try to determine “minimum quality of service” requirements for Internet provision. The Internet’s unprecedented growth and the wide range of unexpected uses it has found are widely attributed to the “end-to-end” principle, which allows users to innovate by making whatever use of IP packets they can think of without needing the permission of the network operator. The “network effect”, where services become more valuable the more people use them (pity the person who owned the very first telephone in the world!) also helps successful services to grow rapidly. Most Governments agree that the Internet model of innovation needs to continue, but this requires there to be a certain amount of bandwidth and a certain number of end-to-end connected users available both to do the innovation and to adopt the innovative services to the point where the network effect kicks in. Unfortunately it’s very hard to predict what either of those numbers is. Simple consumer preferences (which the Commission are, for now, relying on to keep traffic management at an acceptable level) are very unlikely to protect this generative function of the Internet (Jonathan Zittrain’s term), as that would require customers to choose to buy Internet access services because they will allow them to use services that no one has invented yet! Network providers might have an interest in allowing innovation, provided they see it as benefitting their commercial interests, but if those interests result in traffic management expanding to cover more users and more networks then it is easy to see how the the generative Internet might be squeezed out. How we detect when this is likely to happen, and how we stop it happening, seems to me the most interesting part of the network neutrality debate.

Out-law have a summary of the current legal position on traffic management and network neutrality.