Fixed Access Networks in Berlin

Better bandwidth in Europe and the world - how depends mostly on where.

I attended the Fixed Access Networks Summit last month in Berlin and chaired day 2.

A very useful and in my experience one of the best conferences I’ve been at for a while.  Credit has to go to Informa and the conference organisers, led by Anja Profozic, for putting it together. But it was the quality of the speakers and the current uncertainties in a number of areas that made it particularly interesting and timely.

The primary message is that the solutions required in terms of fixed access broadband provision vary according to the local conditions of a market or even down to the individual property level at times.

This doesn’t make it any easier to satisfy those requirements (or even at this stage fully understand them) but it is a sign of the shift in the mentality in the industry that the confrontational element of fibre versus copper, in particular, has settled back a little.

There is the recognition that hybrid networks will exist for some considerable time. Although, where there is legacy infrastructure and the acknowledgement that FTTH/P is the logical solution greenfield solutions is now the norm.

The cable versus other fixed solutions still exists of course.  Exacerbated by the dominance of Liberty Global in the HFC market, these traditional bandwidth competitors are heading for “Bandwidth round 3 – the gigabit tiers”.

 

  • FTTH/P – Skanova will supply a household in Sweden with a fibre line, and allow them to choose their own ISP, providing there is a one off connection fee of up to €3k paid by the home owner. The model has worked very well as the ‘education’ of the consumer has proceeded more quickly than in some markets;  the message that a fibre line adds €5k-€6k to the value of a property has been part of the toolkit in persuading property owners to partially fund the deployment of FTTH.Bangladesh (Bangla Phone) have also had some success in persuading their new customers to pay one off connection fees.  Although, not at the same level as Skanova but still a significant amount particularly in the context of their incomes.This hasn’t worked well in other markets yet, see FoD in the UK, but perhaps with the right conditions it can be used as a resource that some places have yet to tap.

    The trigger point for a switch from hybrid copper to end-to-end fibre will vary from market to market again.  You have lower OpEx with a mono-tech network but the CapEx to get there is still too high for many.  As fibre gets closer and closer to the customer premises the calculations will be run again and again.

    There was also discussion on the role of FTTH in rural settings, where the low cost of overhead deployment and the efficiencies demonstrated by BT in Cornwall add more grist to the analytical mill; when looking at the composition of the areas that you want to cover.

  • Vectoring – which is a necessary part of offering significant bandwidth uplift on the next generations of fibre/copper networks. It is however very difficult to do if you need to be able to unbundle lines at the cabinet and sub-cabinet level.Italy and Fastweb have addressed this issue using a sophisticated but potentially complicated (and expensive?) hardware solution that will allow suppliers to share cabinets.Germany have adopted a first come first to serve (the ISP ‘owns’ the cabinet) methodology, while others are touting the application of SDN enhancements as the way forward.

    Other markets have just removed the requirement from the incumbent to offer anything other than VULA and bitstream solutions.

    Therefore, we will end up with quite a patchwork throughout Europe in the next five years which could be a problem, particularly in the context of any potential single digital market.  It will also make each solution more expensive as volume pricing for kit in particular is compromised.

    The whole topic of vectoring and far end cross talk (FEXT) will continue to be part of various solutions and discussions in the next couple of years.  See the next point.

  • G.fast and Vplus (Supervector etc.) – both these solutions are going to be appearing in mass deployment in Europe from 2016 although possibly later for Vplus.G.fast is now offering more bandwidth further out bringing many customers within reach of realistic cabinet deployment (as opposed to FTTS or FTTdp) options and bandwidth uplift. The primary blocker at the moment appears to be the potential port density.  As yet 8 and 16 port units only are actually available in reality although not final production models and the steps up to significant port density are still to be taken.Vplus (or 35b) offers less bandwidth than G.fast closer in but does better at the 1000m and above ranges.  However, the 35b refers to the frequency that Vplus will push up to, instead of 17MHz for plain old VDSL.  This interferes with any potential upgrade path from Vplus to G.fast that requires the full 20Mhz-100Mhz range to function best.  So although it is being touted as a stepping stone to higher bandwidths and as ‘complementary’ to G.fast it doesn’t look particularly efficient to do one then the other.

    So at the moment Vplus will allow more bandwidth further out and is backward compatible with VDSL and allows the CPE to remain in place but isn’t standardised (the timetable for that isn’t certain). Furthermore, this won’t make Gbps tiers available and will also push any G.fast deployment further down the road.

    On the other hand, G.fast is getting more bandwidth to higher distances but does need changes at the CPE and cabinets; as well as needing work on port density before cabinet deployments become straightforward.

    What works best and where will be subject to a number of calculations.  Although, cost will feature heavily so will the market context, politics, decisions already taken as well as the subjective views currently around for future demand.

  • Cable operators – DOCSIS3.1 will be appearing from 2016 too. Partly as part of their upgrade cycle but also in response to the competition which cable operators have for so long being responsible for driving, at least in bandwidth terms.The speakers laid out their understanding of what conditions may or may not sway the outcome of cable v G.fast in particular.  The current Achilles heel of cable is the upstream bandwidth availability and if applications or demand emerges (it’s not clear it will) for symmetric supply in a residential environment it would be advantage FTTx.

 

The most fundamental uncertainty at the moment is what bandwidth will residential users demand in five to ten years’ time?

We have our own view and it hasn’t really changed, Bandwidth for Video: How much is enough It is entirely possible from evidence that has started to feed back, in terms of gigabit take-up in FTTH markets and the pricing pressure on 100Mbps and above services, that there is very strong demand for superfast and ultimately ultrafast services.  However, the message we hear at the moment is that mass residential adoption for Gigabit Tiers will start to kick in from 2020 onwards, if not even further in the future.

Please follow and like us: