European Consultation on net neutrality - draft answers Part 2

Question 11: What instances could trigger intervention by national regulatory authorities in setting minimum quality of service requirements on an undertaking or undertakings providing public communications services?

There are several answers to this question, which we can separate into contraventions by ISPs of what I term net neutrality ‘lite’ and ‘heavy’.

‘Lite’:

1. Blocking of Skype/VOIP products – this is especially true of mobile networks which have made no secret of fairly widespread blocking. See Madison River (2005) FCC compliance order.
2. Blocking of P2P – when protocols or programmes are blocked rather than a general throttling for peaktime traffic management, this is directly discriminatory. See Comcast (2008) FCC compliance order.
3. Blocking or throttling of video traffic whether delivered by CDN (Akamai or YouTube for instance) or P2P or streaming – which is also discriminatory as in Case 1. See BBC iPlayer/BT (2009), Turkey/YouTube (ongoing) or Norwegian case (2006) for examples.
4. Blocking or throttling of more general traffic classes at hand-over point between ISPs – see AT&T/SouthWesternBell [all now merged under brand ‘AT&T’] (2006) merger consent order.
5. Note all these cases represent blocking and throttling at either end-user or hand-over points. It is significantly harder to analyze hand-over point discrimination for affected parties.

‘Heavy’

1.      The questions here concern the nature of ‘unreasonable’ delays in upgrading non-premium (i.e. 99.999% of Internet ) traffic.

2.     It should be de facto unacceptable ISP behaviour for throttling to be introduced for more than a minor part of each day – perhaps 6 hours. Many ISPs now throttle for much of the business day and the evening peak with no break, e.g. 10am-11pm with a 2-hour mid-afternoon break (Virgin Media terms of May 2010).

3.      It is also clear that the consultation on Universal Service needs to make very clear that this universality must be to open Internet access, not throttled access to premium services first, and Internet second.

4.     To discourage continued ISP under-provisioning of bandwidth requires two strengthened regulatory interventions:

[a] to prevent misleading advertising and mis-selling based on that deceptive advertising, not simply within generally weak consumer protection law and even weaker advertising self-regulation, but also in General Conditions of ISP licences;

[b] to remove bandwidth bottlenecks which are generally no longer in the local loop (where ADSL2 is increasingly common at least in urban and semi-urban exchanges, and DSLAM cost is decreasing as Moore’s law drives down processing costs). This indicates increased NRA concentration on backhaul costs which are still typically the monopoly or near-monopoly of the incumbent with SMP. Both duct-sharing and functional separation can eventually act as incentives to drive down costs in a competitive market, as well as Ethernet LAN products and other NGA backhaul products, but NRAs appear spectacularly misinformed or misguided as to the exponentially decreasing costs of providing such backhaul, instead allowing themselves to be sidetracked by incumbent pension fund deficits and other issues.

The Commission must make plan in its forthcoming NGA Communication that the cost of backhaul remains the key cost for ISPs and it is this that drives short-termist throttling strategies rather than longer-term investment in bandwidth.

Question 12: How should quality of service requirements be determined, and how could they be monitored?

One of the more effective actions by Ofcom has been its joint exercises in testing Internet speeds with Samknows and this should be examined at EU level. However, as a lawyer rather than network engineer, I defer to those with specialist knowledge in this area. I caution that a body of technical experts to determine such factors must always be co-regulatory rather than self-regulatory (as the US approach appears increasingly).

Question 13: In the case where NRAs find it necessary to intervene to impose minimum

quality of service requirements, what form should they take, and to what extent should there be co-operation between NRAs to arrive at a common approach?

Quality of Service minimum levels are the only way for independent application and service developers, and their financiers, to be confident that their innovations can be received by consumers. A single European market in Internet innovation depends on a common minimum QoS standard. That is surely obvious to all. Failure to create a single minimum – above which NRAs are of course welcome to set higher standards, particularly in higher-speed regions within countries – would jeopardize the single European market and signal to venture capital to look elsewhere for entrepreneurs to support.

Question 14: What should transparency for consumers consist of? Should the standards

currently applied be further improved?

The means to achieve this is a consumerist intervention to prevent ISPs advertising ‘Internet access’ when they actually provide ‘special access’ to premium services.

A further means to provide true consumer transparency is to force ISPs to advertise not their ludicrous laboratory-feasible maximum speeds, but their true and provisioned MINIMUM speeds.

If they have 24mbps backhaul and service a maximum 800 customers from one exchange, then they should advertise ‘minimum bandwidth’ as 24576/800 = 30.7kbps.

Even here, they must caution that the 24mbps is a theoretical maximum and not all customers can access 30kbps at peak times.

An example should be included – “at peak times, you may not always be able to access web-based email”.

Question 15: Besides the traffic management issues discussed above, are there any other

concerns affecting freedom of expression, media pluralism and cultural diversity on the

internet? If so, what further measures would be needed to safeguard those values?

Traffic management is censorship at its most basic, whether conducted by private or public bodies. This does not mean it is necessarily or always bad – filtering spam for customers has been the default position for most ISPs for half a dozen years. However it clearly affects fundamental rights to privacy (consider the PHORM experiment in behavioural advertising using Deep Packet Inspection) and freedom of expression (including the right to receive information via the protocol of one’s choosing – e.g. via P2P or streaming).

The monograph I published earlier this year considers the wider details in more detail in its final Chapter 8. See:

Marsden (2010) Net Neutrality: Towards a Co-regulatory Solution, Bloomsbury Academic: London.

See also my latest article on the subject:

Marsden, Christopher T. (June 8, 2010) Network Neutrality and Internet Service Provider Liability Regulation: Are the Wise Monkeys of Cyberspace Becoming Stupid? Available at SSRN: http://ssrn.com/abstract=1622324

Comments welcomed?!