CEO.digital CEO spotlight: David Hilliard, Mentor Europe
David is the Chief Executive of Mentor. He is a trusted hard-core specialist with encyclopaedic skills on strategy execution – a down-to-earth practitioner, with over 30 years’ experience in telecoms, media, and service industries.
He has helped many clients setup new businesses – Energis, Eircell, Virgin Mobile, Three, 5 Television and Tele2. Other key clients include Airwave, BT, Cable & Wireless, Dell, eir, EE, Lilly, O2, lastminute.com, Vodafone and BSkyB.
Challenges of future network densification
In 5 to 10-years’ time, it won’t make practical or economic sense to run 5G networks using standard BT EAD/LA ‘lit’ services over a hub-and-spoke architecture.
For heavily densified networks, EAD/LA circuits will be too expensive for the mobile operators. The £2/4k annual price of a BT 1G/10G EAD/LA service – returning to a BT Exchange – doesn’t compare well with a short dig to add a small cell to an existing fibre ring.
What’s more, future price declines in 10G/1G services will certainly be offset by a need to move to 100G/10G.
With an all-purpose ‘lit’ solution, a hub-and-spoke architecture is fundamentally inefficient and inflexible. Especially, where Macro – Micro aggregation must be routed, via the serving exchange.
Worse, there’s no resilience and few options to innovate at the optical level.
Hub-and-spoke architectures also create timing/synchronisation challenges, especially for TDD spectrum and challenging CRAN deployments. Both demand high bandwidth and low latency.
5G mobile will need massive amounts of fibre
A fibre ring is a basic necessity for the mobile operators. A ring that connects macros, supports 5G architectures and solutions, and allows for simple, low cost, on demand, densification
Even if the Small Cell Forum and FCC forecasts of up to 350 Small Cells per km2 are a tad optimistic, lower deployment levels (50 to 100 per km2) will require huge densification very soon. And this will completely blow the economics of an EAD-led architecture.
But, today the price of EAD/LA 10G, makes the case for building a new fibre ring questionable. As things stand, it would be less trouble to use EAD for each new site – or to upgrade capacity to an existing site.
So, the critical question is ‘will the vice like grip of a BT hub-and-spoke EAD solution be broken?’
But when, how and by whom?
Lessons from history
It’s useful to compare where we are today with the history of FTTH in the UK.
In 2012/13, BT – and others – proclaimed there was no business case for FTTH. Apart from ‘new build’ and a very small subset of the population, VDSL and G.Fast would provide all the speed any home could ever need.
By 2017, at 2% FTTH penetration, the UK lagged all international benchmarks so that particular fairy story disappeared without trace. There’s now widespread acceptance by the industry and government that FTTH is the only way forward – and the race is now on.
But guess what happened in the meantime?
BT bought itself 5 priceless years, where it’s been able to use its cashflow to manage the conflicting demands of Pension deficits, Dividends, Football rights and Network Capex.
Is it possible that BT hoodwinked us all?
The shift from EAD/LA will be driven by a combination of aggregated Mobile Operator demand, densification needs – and economies of scope.
Existing Alt.Net fibre and ducts don’t help to ease new build costs
… even in Central London. That’s because Operator demand is concentrated along the busiest streets and ‘hot spots.’
It’s understandable why Alt.Nets have used secondary streets instead. These are the most expensive and time-consuming routes to dig.
At best, this means existing Alt.Net ducts, (if they’re not full), could provide around 30% of an MNO fibre network, in Dense Urban areas only. But the costs of integrating new build, with undersized sections of existing duct would dribble away most of the ‘cost savings.’
With 100% new build, only BT PIA ducts offer the potential to significantly drive down the costs – assuming two things.
First, it assumes PIA is as straightforward to use as many of the “evangelists” suppose. And, second, that the regulatory restrictions on its use are eased.
Operators simply don’t want to buy, light and run their own dark fibre network – unless they’re shoved into it. But they will experiment with self-deployment and trials until they persuade the supply industry to unbundle prices and rebuild various network elements to mirror the main cost drivers.
Today’s capacity driven price model is not lined up with the real drivers of cost. Not by a long shot. 80-90% of fibre build cost is wrapped up in the ‘civils’ and not linked to capacity. The ‘lit’ overlay is a trivial part of the overall cost.
Next steps in deployment
Recent spectrum auctions and MIMO solutions buy a little time to get solutions in place – but they don’t “solve” the Operators problem.
Given where they are now, UK Operators face a 3 to 5-year journey to 5G networks – built on full fibre-ring backhaul – and densification at scale.
As with FTTH, the UK will lag behind deployment speeds in many other markets. It will take time to aggregate demand and deploy the crucial building blocks.
But we will see lots of trials to test different technical and commercial constructs.
There is nothing new in physical deployment.
Timescales will be driven by the scale of deployment – and the complexity of aggregating demand and requirements across multiple parties. This is where the economies of scale and scope will come from.
But even with a sluggish start, there is some sunlight on the horizon.
Deployment is bound to start in Central London soon – then spread to Urban London, followed by the Centre of other major UK cities.
And that is very good news.