Exploiting network densification to accelerate 5G deployment
Mentor were delighted to get involved at the Connected Britain conference, held in London on the 14th and 15th June 2017.
The conference was opened by Matt Hancock, the Minister of State for Digital and Culture and fielded other stakeholders and speakers from the telecom industry involved in delivering high speed connectivity in the UK. These included senior representatives from operators, altnets, vendors, market analysts and the investment community as well as central and local government.
It was a well-timed event focusing on the challenges the UK faces in becoming a vibrant and robust digital Britain against a backdrop of political and economic uncertainty.
David Hilliard, Mentor Founder and CEO, delivered a candid and provocative speech focusing on network densification and the UK’s limited country-wide access to competitively available dark fibre.
Read the full transcript of David’s speech below.
Written and delivered by David Hilliard
I’ll start with a simple statement.
If we don’t decouple transmission costs from traffic volume – then 4G won’t deliver its true potential – and 5G is dead in the water!
Let me explain.
We all understand Mobile is at least as important to a “Connected Britain” – as Fixed networks are.
But the mobile industry is in a bind; saturated markets and a generally flat revenue picture.
Traffic growth demands more and more CAPEX – but offers smaller and much-tougher-to-achieve returns.
Many people today talk about market discontinuities. Well, here’s mine.
The insatiable demand for capacity – driven by smart phones – is shattering the normal capacity growth strategies of Mobile – and forcing operators to dive into network densification.
For those that haven’t heard about densification – my guess is there won’t be many – it’s the process where an operator massively increases the number of base stations in high demand areas.
Alternatives to Densification
You might well ask – aren’t there other strategies to increase capacity?
Well, there are – several.
- Wi-Fi – for one – has been the gift that keeps giving. Off-loading masses of data traffic from the Mobile Networks. Analysis Mason has claimed Wi-Fi carries c80% of smart phone traffic. But in dense urban areas, where mobile will exhaust first, Wi-Fi doesn’t seem to have much more to give.
- Smart Antennas and other techniques – to optimise existing spectrum on today’s base stations are useful – but they’re incremental step-ups – rather than game-changers. With 40% annual traffic growth, they don’t buy us much more time either.
- New Spectrum is a very attractive solution – but they aren’t making any more any time soon. The Spectrum that’s left is at higher and higher frequencies – which shrinks the cell size anyway – and may not be supported on older phones. We can re-farm old GSM and 3G spectrum – but we’d need to switch off these networks and services – and you know how good we are at that!
But back to Network Densification – which I think is the way forward for the industry.
This is the process where we times 5 – or even times 10 – the number of base stations in a given area – and reuse the existing spectrum many times over. Done right, this promises a radical increase in network capacity.
But it’s not a stroll in the park. We need to manage the user experience very carefully – dealing with interference between small cells – by actively using lower radio power – careful planning – and special software.
This is the essence of 5G – and a big part of our future anyway.
Urban Reasoning for Dark Fibre
We can’t just deploy thousands of new base stations. It’s just not affordable
O2’s recent announcement of its intention to roll-out 1,400 “small-cells” across London – to boost mobile phone reception and prepare the capital for 5G – shows this is not blue-sky stuff. It’s really happening.
There are some snags, however.
We can’t just deploy thousands of new base stations – if the costs remain the same – as for the current-style base stations. It’s just not affordable. We need all of the costs of the small cell to reduce. Site rentals, the kit and the backhaul.
Don’t be fooled by the word small. Small cells are actually very high capacity nodes in their own right – and each will need its own high speed backhaul connection. What’s more, the technology that allows us to manage interference – and the user experience – needs very high-speed fibre connectivity back to a “control node” that might only be a few blocks away.
Some of the leading technologies to solve this problem specify dark fibre – for these short but very demanding connections.
Pressure to decouple transmission costs from traffic volume is not going away.
Let’s be very clear. These densification plans are tough, complex programmes to execute.
But it’s about WHEN not IF – in my view.
The Market Environment – a Fragmented Picture
Gerry McQuade who runs BT Wholesale recently spoke at the Wireless Broadband Congress in London. He poured cold water on densification by saying …. ”the economics don’t work to go from 17k sites to 50k+ sites!”
Well, I think he’s right for TWO reasons.
- Firstly, BT/EE has so much spectrum – why would you be first to densify when you can just add more Spectrum from your stock? Obviously, this puts off the fateful day for BT/EE for several years.
- Secondly – and the heart of my pitch today – the cost of transmission is too high and too closely linked to traffic – to make densification affordable. BT Wholesale and Openreach dominate the fibre mobile backhaul market in the UK. And they know how much this stuff costs.
So, pausing for a moment – we have seen that mobile is at the heart of “Connected Britain” – and that mobile traffic is rising faster than on fixed broadband.
4G has been a great success delivering on the mobile data promise – and it’s worked hand-in-hand with Wi-Fi.
The functional 4G roadmap is clear – and most of it is now deployed. This means that operators with spectrum challenges on the horizon must now move quickly to densify their networks – in an affordable way.
And dark fibre – is at the heart of this densification process.
So what’s the problem …… ”let’s just bring on Dark Fibre!”
We don’t have an abundant supply of Dark Fibre in the UK today.
A typical UK Mobile network has c17k base stations. Let’s say half are in urban areas – where extra capacity will be needed first.
Densification could drive this number up significantly – maybe over 50k sites in 5 years. This is a big backhaul challenge – and opportunity. Especially when you multiply by 4 for all MNOs.
One estimate quoted at the Small Cells Forum in May 2017, suggested in excess of 500k small cells for London alone – between 2020 and 2025.
Of course, Dark fibre is readily available for intercity/inter data centre use – but these new base stations need metro and local fibre solutions in busy urban areas.
The Alt-nets do a great job promoting dark fibre products in their metro patches – where they have built out. But that’s not everywhere.
Overall, it’s a fragmented picture.
Lack of Competition
BT Openreach is the dominant provider of Local Fibre in the UK. It has lots of fibre just about everywhere. Openreach doesn’t offer dark fibre today. But, nudged along by Ofcom, they are launching a Dark Fibre portfolio this October.
So, we like regulation, but we like competition more. It drives businesses to improve, respond and invest.
Incumbent Telcos are tough businesses to compete with
They hold so many of the cards. But we have seen them react.
Take Google Fibre. Their Fibre threats in the US seems to have focussed minds. As a result, local Telcos there are responding with their own products and investments. So far, Google hasn’t actually built much of its own fibre. It’s been done for them.
Likewise, here in the UK – BT raised its game on broadband when Virgin offered higher and higher speeds – and they really responded when Virgin announced their new footprint investment.
The question for us is:
- Will the market set-up we have today do the job we want with Dark Fibre?
- Will it have the right pricing – the right service wrap?
- Will it be good enough – and quick enough – for the Mobile operators to make this jump to densification?
Without a real competitive threat, I think the answer has to be no.
Openreach is being pushed to do dark fibre
We can’t ignore Openreach – but almost everybody finds discussion about the business uncomfortable.
Look, they’re good guys. I respect their discipline and delivery capability – but Openreach is being pushed to do dark fibre. And what we actually need is an enthusiastic and committed supplier of dark fibre.
We don’t yet know the final product detail – but BT probably sees this as lost or ‘at risk’ margin on managed services, rather than a whole new market for dark fibre. Some competitors may see it as a chance to unbundle Openreach – and somehow steal margin and value – by lighting up fibre to nurture their own managed services.
If this turns into an industry brawl, we won’t make progress any time soon
What we need is a collaboration – with enough missionary zeal – to deliver new dark fibre for dense Mobile.
One example to illustrate the problem is the repair time. Dark fibre is a bit like an unbundled Ethernet service. Repair times for BT Ethernet is 5 hours. Yet, the proposed repair time for the BT Dark Fibre is 18 hours.
You can just see what the BT argument might be. “We don’t provide the active electronics, so we can’t detect faults and repair them as fast as in Ethernet – where WE light the fibre.” An eager dark fibre supplier would say – ”how can we work together with our customers to achieve the same 5 hours with dark fibre that we manage to achieve with Ethernet?”
We’ve seen BT resist Dark Fibre for many years.
It’s not because they are bad people. They are the dominant player – and many companies would do exactly the same in their shoes. Why rush to risk margin – and value – in their massive installed base? It’s much easier to move backwards – slowly – resisting Ofcom in a skilful way.
Let’s talk about Ofcom for a minute.
We’ve all seen Ofcom cracking the whip over the last while, threatening to break up BT. It’s hard to see what more they could do now. I’m sure they are close to the limit of their powers. So, BT are behaving entirely rationally – and Ofcom is a top-notch regulator.
But we still don’t have dark fibre at the right price – and in the right places for densification.
Dark Fibre Supply Options
Our train is hurtling toward the densification discontinuity – but quite likely to swing off onto the slow track, due to the scarcity of Dark Fibre
I really believe this is a critical moment in time for the industry – and it’s going to happen very soon!
So how can we fix this?
There are only three options to get the right UK dark fibre supply, in my view:
1. Wait for Openreach to become a great dark fibre supplier – this will happen but it will simply take too long. There’ll be no rush to drop prices in the short term and the risk is mobile capacity build will stutter and fumble while we wait. You may have heard Clive Selley, Head of Openreach, ask the industry to meet him and discuss investing in fibre and broadband. That’s great – but won’t we just end up funding an extension of Openreach’s monopoly?
2. We could create a national dark fibre alternative to Openreach from scratch – this is probably unrealistic and massively expensive. It would take too long and be too risky. And anyway, why “overbuild” the alternative fibre investments already made.
3. Build a collaborative solution between Dark Fibre users and existing Providers – to drive investment; eliminate fragmentation; fill gaps – and create a credible alternative national supply for mobile operators. Sort of crowd-sourcing an alternative to Openreach.
But where do we begin?
Dark Fibre Alliance
Looking at the business case would be a good place to start
We’ve done a detailed piece of work to see if there is a case to invest to solve the Mobile Dark Fibre problem. Let’s just look at some of the critical factors and what they tell us about a possible solution.
First of all, let’s look at the BT Dark Fibre pricing. Good news – there is reference pricing on BT’s web site – but it’s not clear if this is the final price for launch. In any event, dual fibre is dearer than single; two ends pricier than one; and, out of area more costly than own exchange.
The 20 year TCO is over 95% OPEX and there is no tapering of price, based on volume or geography. Even resilience seems to be priced as a second full circuit.
We’ve looked at the cost to dig fresh fibre infrastructure. It’s not cheap or easy – but the 20 Year TCO is over 70% capex. What’s more, if you dig, you can put as many fibres in as you like. It doesn’t move the costs much. And by carefully planning the network in rings – resilience can be almost free.
Now turning to demand from Mobile Operators
We’ve split the country into different “Mobile Geo Types” – Rural, Urban, and Dense Urban.
It’s clear that if you build your own fat dark fibre network – it only works if you can win enough business to achieve a payback. That suggests Dense Urban is more attractive than Rural. It’s not difficult to understand. But it also says that two operators – or more – are better than one. And operators that are densifying will be better off than those who don’t have plans – just yet.
Digging in Dense Urban is not a walk in the park.
But we know that alternative fibre infrastructure already exists. So, let’s assume we use it – where we can – rather than overbuild it. We need to buy it on an IRU basis – capex rather than opex. We think those deals can be done.
We also need to explore alternatives to digging up our hard-pressed roads and pavements:
Alternative ducts – like those being installed for electric car charging points – and water assets – are well worth exploring.
What about point-to-point radio? Can that help us avoid burying fibre at all – in some places?
It’s easy to imagine a practical mix of solutions – giving some upside – instead of a new traditional dig which would improve the average costs we would face.
Bottom line is – this could work with the right coalition of providers and users committing to buy – for the long term.
The fly in the ointment is these key players are all competing with each other in the market place today – so it’s not straight forward. However, thinking about the business case and benefits, it is a really exciting prospect.
But, there are some open questions:
- How can we improve coverage and fill the gaps?
- How do we trigger new build – if required?
- What about common standards?
- Can we reduce the integration costs for users?
We can’t expect MNOs to consume Dark Fibre from dozens of different suppliers – using different interfaces and processes. Can MNO’s put it all back together – and achieve the same service levels as from BT? Can we add resilience by linking different metro networks?
We want better service – not worse.
So, for network densification to work economically, it’s clear we must disconnect transmission costs from growing traffic volumes.
We have no choice.
Why would we sleepwalk into a situation where we have no alternative to Openreach?
BT Openreach is unlikely to offer much useful help in the short term. So, why would the rest of us sleepwalk into a situation where the industry has no practical alternative to Openreach.
There are alternatives – but we need to grasp the nettle now. And the harder we grasp it, the less it will hurt. There are people sitting in this audience today who own key parts of what it would take to crack this problem.
- have dark fibre
- want to buy dark fibre
- have skills or technology to help exploit it or may even have ducts we could thread fibre through
Or maybe you just share my eagerness to fix this problem?
If you do, please get in touch. I’d be happy to share more of our thinking and hear your story.
Who knows – perhaps we can join forces and fix this – and make a bit of history in the process.