FTTH and the information super mud track

The following questions were circulated to speakers on my panel at the FTTH Council conference. I’ve tidied up my notes (assembled with much help from my friends) and reproduced them here.

Q1. Do you believe the UK and EU policy is going in the right direction and striking the right kind of balance between investment, competition, innovation, choice etc.?

No, we are not creating an environment for rich service competition together with strong economies of scale. The kinds of wholesale products that ought to be emerging are notable by their absence.

Indeed, we’re going backwards. Whilst people talked about the “information superhighway”, what we’re building is the “information super mud track”, where everyone is bogged down in contention. We’ll carve large numbers of parallel tracks through the landscape, and the journey times over them are going to be as unpredictable as transit times over mud tracks in the ever-changing British weather. The very concept of metalling the roads (i.e. actively managing contention to allow smooth passage) has yet to become accepted normal practise.

The worst part is that they are all some kind of pay road with an annual subscription, not even a toll road where you pay for the outcome of a successful crossing.  Once you’ve chosen your pay road provider, then you are stuck with them. There is no timely redress if your traffic gets stuck in the mud. Your only sanction is to move to another mud-spattered pay road system and argue over your remaining annual charge.

Q2. How do you view FTTC, will it enable or impede the deployment of FTTH networks?

Impede. It gives fibre a bad name, delivering weak performance due to poor flow isolation. Furthermore, it continues to play into the common myth that more edge capacity assures better fitness for purpose. It doesn’t – the extra multiplexing point in the street cabinet impairs performance for latency-sensitive applications.

Q3. How do you view the European Commission’s initiatives (to be implemented by UK Government) to reduce deployment costs through duct, trench, and pole sharing etc.? Do you see new investors entering the market? Who are these cost reduction measures aimed at in your view? How important is public funding going to be?

Adding in more layers sounds great in theory, until things go wrong: see the railway industry for the blame-passing that results. We need a national infrastructure with a common interface for services like M2M, smart grids, e-health. Public funding is fine for passive infrastructure – but the focus should be on raising the benefits, not paring the costs. The fragmentation of the network into muddy fields destroys the ability to innovate, as the ability to create cohesive dispersed communities of interest at low marginal cost disappears.

Q4. What role do you see for more profound policy interventions such as structural separation where the network assets are held in a really separate company (so in the UK, Openreach would be owned by a different group of people from BT)?

There are pros and cons.

In favour, it would certainly create a different set of incentives, and enforce greater transparency. In principle it represents progress, if the right kind of active and passive wholesale services are enabled and encouraged.

However in practise, creating more boundaries doesn’t represent the answer. The problem is the new services require economies of scale. The innovative steps (such as smart grids, small cells or home healthcare) find the mud tracks to market proliferate in all directions. They are so diffuse and unpredictable that the application investment is never undertaken.

In theory you can create standards so everything works around a common interface, but that theory hasn’t really worked at any time since SS7 and GSM were deployed. The distributed computing world is now too complex. Maybe one day it will modularise, but not yet.

Q5.  Are a series of local FTTH builds likely to undermine the benefits economies of scale and lead to a patchwork quilt of networks across the UK? Should we be worried or can the economies of scale still be realised?

Yes, they are undermining those economies of scale, and we should be worried. Many of the wider benefits will be missed as a result. Markets are poor at infrastructure, and relatively weak at social services. You need to be able to compose standard service components, using clean and common interfaces. Until then, they are a costly way of turning a broadband bog into high speed mud tracks, not a common highway system.