How to upgrade and deploy communications infrastructure for 5G

Overlaying 5G radio equipment at the ~20,000 cell sites operated by a typical MNO in the UK, Italy or France would cost an estimated US$1.25bn per operator, per country, and that’s before one considers the costs of densification, indoor coverage, edge computing and upgrading the core network. The telecommunications industry will need to access long-term, low cost capital which, combined with the impetus toward shared infrastructure to lower economic thresholds for 5G deployment, suggests deepening partnerships with towercos. This article combines insights learned from panel sessions featuring Arqiva, Cellnex, WIG and MBNL, with round tables led by Cornerstone and EdgeInfra at the TowerXchange Meetup Europe 2019.

“We’re investing now in anticipation of 5G,” said Alistair Davidson, a Director at Wireless Infrastructure Group (WIG), one of the UK’s leading independent towercos with 2,200 sites across the UK, Ireland and The Netherlands. “Everything we’re deploying now, from the small cell network we’re building along 35km of road in the West Midlands as a CAV (Connected and Autonomous Vehicles) test bed, to our bid to provide coverage on the London underground – everything is connected by fibre. We’re ensuring short distances to equipment rooms for low latency, we’re using low band antenna systems that will run from 1,700MHz through to the upcoming 3.6GHz band. We’re looking at a new network topology and radio plan, based on 3.5GHz and even millimetre wave spectrum. 5G handsets are in the market already, and it may be just 18 months until there is a consumer proposition for 5G.”

Pat Coxen, Managing Director of MBNL, an infrastructure sharing joint venture between EE and Three, added: “upgrades for 5G are already in-flight in terms of the infrastructure upgrades: the technology infrastructure, cloud cores, transmission infrastructure, and digital infrastructure. Then there is the internal systems component: for example, Three UK are talking about significant upgrades and digital transformation in IT, BSS and OSS systems. And then we have upgrades to the physical infrastructure – activity to upgrade towers and the utility services which support them have kicked off now. Both MBNL’s shareholders are live with 5G trials and say they’ll have 5G services (or 4G++) live somewhere between Q2 and Q3 in the UK. I’m anticipating a hard and fast rollout to the existing macro network, whereas there is less evidence of investment in small cell capacity densification and coverage.”

Oscar Pallarols, Global Commercial Director of Europe’s leading independent towerco Cellnex Telecom, which has 23,652 sites across six countries, added: “From our point of view, there are four principle considerations in 5G deployment:

1. 5G radio colocation on current sites, accommodating the incremental weight and power consumption of 5G, and adapting to customer demands about the positioning of massive MIMO on the tower.

2. Connecting key network elements with fibre: we must deploy fibre to the tower, and to the antenna, in a very short period of time.

3. In 5G pioneer countries such as France, we are already closing some agreements to build central and metropolitan offices to distribute the core network to regional groups.

4. It’s pre-commercial, but we’re seeking to understand whether towercos can create business opportunities in edge computing – we’re developing our thinking on the way to deploy for OTT players on the edge at our cell towers.”

“Much of the 5G deployment will be carried out by, and is already being carried out, by neutral hosts,” concluded WIG’s Alistair Davidson. “We’re having a lot of discussions around the economic challenges of achieving the last 10% of land mass coverage, and the densification needed for 5G, so the neutral host model is gaining traction to unlock multi-operator economics.”

Rhys Phillip, CEO of Cornerstone (formerly CTIL) hosted the most popular round table breakout at the TowerXchange Meetup Europe 2019, focusing on upgrading cell sites for 5G. It should be noted that Cornerstone were one of several infrastructure companies and MNOs participating in the round table session, which was held under the Chatham House Rule, hence the following remarks are not attributed specifically to Cornerstone, nor to any other participant.

The process of upgrading sites for 5G has started in many European countries already. “We have been planning for 18 months, and we’re putting spades in the ground, putting new towers up, and upgrading the infrastructure now”, said one towerco. “We have a long programme of work to get all of the infrastructure ready, but the money’s going out the door, stuff’s being built now.”

“We will be upgrading thousands of sites,” added a tower owner. “A lot of those sites needed to be upgraded anyway, but there’s still a lot of second guessing going on as to how we plan those upgrades as multi-band Massive MIMO weights 80-100kg. If you want to stick six of those at the top of a tower, you probably need a significant rebuild. If you want 12 on a rooftop, you need to rebuild or split that site. We need to find a capital efficient way to do this so when the kit decreases in size so we won’t have wasted money.”

The prevailing view at the round table was that a significant majority street works would need some sort of enhancement as many currently lack structural capacity to overlay existing radio equipment with 5G. Different towercos and MNOs estimated a broad range of proportions of their macro tower networks which would need structural enhancement for 5G, between a low of 0% (“our towers are relatively new, and designed for overlays and co-location”) to 60% (“more than half our macros will need upgrading because of the physical size of antennas and RRUs, and due to the incremental power requirement of 5G in addition to 2G, 3G and 4G antenna”). There were mixed views on the timeline for the potential consolidation of 2G, 3G and 4G into multiband antennas: for some operators, this seemed likely in the near term, for others, it seemed likely that the costs sunk into 2G, 3G and 4G antenna, and the long term leases signed to mount those antennae, meant they might not be consolidated at all.

The prevailing view was that there was a need for an increased level of transparency from MNOs about their 5G deployment plans. “If they take three or four years to devise their strategy, then give us one year to execute, that will put huge pressure on the supply chain,” complained one towerco. This “pressure on the supply chain” will create a significant resource pinch: towercos are lean businesses with a finite field force – most towercos outsource the majority of their construction work, and there are often only half a dozen or so proven, competent subcontractors in each country that towercos and MNOs would trust to upgrade existing and build new towers. If 5G is being rolled out concurrently by multiple MNOs, there will be more demand than supply for scarce communications infrastructure construction resources. It should also be noted that recent years have been relatively slow for telecom construction contractors, meaning many workforces have been downsized – worse still, some competent contractors have gone bankrupt.

“Could this create subsequent maintenance issues?” Asked one round table participant. Most felt that after 5G installation, the maintenance regime for passive and active infrastructure would not change significantly. However, maintaining control over assets, and Health and Safety, would remain key. “Overloading towers would be a health and safety disaster – and one which we simply won’t allow to happen,” said one towerco, citing diligence processes to ensure safety.

Several towercos and MNOs are reviewing their access control processes and systems to ensure only the right people have access to the right parts of each site. “We’re trialling an access control system at the moment,” said one towerco. “Our requirements are not just about site access – what’s key for us is ensuring we maintain an accurate asset inventory, and track the relationship between the site visit and what’s changed as a result of that visit. Smart site access is not just about safety and duty of care, it’s also about avoiding undue stress to landlord relationships, and that means having an accurate digital representation of what is on a site to minimise site visits.”

While most macro towers have sections that can be relatively easily replaced to increase structural capacity, the challenge remains that many tower owners don’t yet know how much load is needed. In some cases, this traces back to a need for increased transparency from MNOs on their 5G deployment plans, in other cases the tower owners themselves need to update their asset registers. Auditing all those sites is time consuming, although more than one towerco piloting AI technologies and drones to accelerate site audits.

Of course upgrading the structural capacity of towers is as much about the foundations as it is about the ‘steel in the air’. Documentation on tower foundations is often incomplete, and many European towers were built for the purpose of a single mobile network operator, without a view of potential mast sharing or future monetisation.

The economics of 5G mean that more and more towers in Europe are destined to be monetised and shared. The ongoing valuation arbitrage between towercos and MNOs (wherein the former can be valued at 2-4x the latter) continues to make tower monetisation a compelling financing model for 5G.

Not every tower portfolio will need significant upgrades for 5G. In their panel session, WIG suggested “in terms of the work needed to upgrade structures, I hope less is needed on our towers than most as we invested from the outset. We won’t build a macro tower less than 25m high, and usually build that as a stub tower so we can easily build up to 35 as soon as we get the demand coming through. We have already invested in asset management software, and have re-engineered all our processes so we were ready to do all the upgrade works, and we can see the status of all those works. We have a process and a system so we can upgrade quickly.”

If upgrading the structural capacity of towers to accommodate heavy 5G antenna is one driver of change, another is the increasing power requirement. “We anticipate 5G increasing the power requirement at cell sites by 3-5KW,” said one towerco.

Assuming for a moment that the electricity grid can still provide primary power to cell sites in the 5G era (and that assumption is risky, particularly if EV charging is co-located with many cell sites), are European tower owners thinking about upgrading power and backup power for 5G?

“It’s under consideration yes,” said one tower owner. “We’re not big users of diesel gensets (DGs) but we’re considering upgrading 15KVA DGs to 25KVA. We’re also experimenting with solar. As an industry, we can be inclined to just slap in more batteries to cope with increased load, but smart site management systems should tell us where we can stretch existing assets to the limit of what’s technically possible, and where we need to invest.”

Another tower owner highlighted the increasing strain on air conditioning systems. “There are always trade-offs – cooling is a constant and takes significant space. Cooling is often primarily there to cool batteries, so maybe this will prompt us to use high temperature batteries.”

“Power is inherent to the data centre business, and edge data centre operators must take responsibility for primary and backup power,” said an expert in that particular field. “If cell sites are to host edge data centres, there may be an opportunity to use batteries to stabilise the electricity network and ease peaks in power demand.”

There has always been a bifurcation between developing market towercos, many of which provide power as a service, and developed market towercos, who typically provide only “steel and grass”, while power remains the responsibility of their MNO tenants. Will the 2-3x increase in power requirements for 5G prompt MNOs to force towercos to take responsibility for power? If more and more cell sites are shared, can we justify having parallel sets of backup power systems? And if a cell site is going to be more than a cell site – an EV charging station or an edge data centre – will towercos be forced to provide power as a service in developed markets? There are more questions than answers – for now!

While there might be use cases for 5G in rural pockets (Agritech applications for example), not everywhere will need a 5G site, and urban and dense urban areas will be covered first. That said, if CAV are to be a key use case for 5G, then filling in “not spots” on A and B roads is going to be critical. 5G spectrum licenses may compel MNOs, even in countries with 98%+ population coverage, to expand networks further into the economically challenging last 10% of geographical coverage – however, those rural coverage requirements are likely to be technology agnostic.

One round table participant shared an example from part of a government white spot coverage programme seeking to have 50 rural towers built. Connecting one of those towers with fibre would have cost ~US$2.5mn – and there were 40 inhabitants within range of the tower!

There are a growing number of rural RANsharing programmes in Europe, one example being Cornerstone’s Project Beacon involving Vodafone and Telefonica (O2). There has been talk of 3,000+ new rural sites needed in the UK, 15,000 in France “such new rural sites may only be economic if there is RANsharing,” suggested one commentator.

Alternate site typologies can represent compelling potential rural cell sites. One Dutch participant highlighted the country’s many windmills, where in many cases transmission and power are onsite, as well as the vertical structure.

Towercos don’t believe in the saying “if you build it, they will come” when it comes to rural coverage – if there is no anchor tenant, indeed no clear pathway to a second tenant, the economics of building rural sites simply do not add up. Some towerco business model exemplars show ROI on a single tenant tower as low as 3%, while with a second tenant ROI rises to 23-30%. Zero tenant towers are obviously pure cost, but even single tenant towers suppress towerco margins. Until the capital cost, and operating expenses, of rural sites can be improved or subsidised, towercos are no more motivated to pursue the rural coverage opportunity than MNOs.

“Each nation has a choice,” said one round table participant, “either to leave rural coverage to be commercially led – which will take longer and be patchier – or for the treasury to take money out of licenses and spectrum auctions and make that capital available to subsidise rural coverage.”

Another area for potential government intervention is to address the rising cost of the land under towers. In the UK the Electronic Communications Code was changed at the beginning of 2019 with a goal to ease rollout of 5G – has it worked?

“Not yet,” said one towerco. “The legislation changed gave towercos and MNOs the right to renegotiate with landlords, valuing site leases relative to the value of sites’ alternate use. The changes have also enshrined rights to add more equipment to towers. But since the code came in, landlords have not been rushing to embrace renewal – so it’s temporarily stalled the lease renewal regime in the UK. We’re bringing cases to a specially created Land Tribunal. Those tribunals are taking time but seem to be heading toward fair resolutions – and those will define principles and precedents. So will it unblock the current hiatus? Yes we hope so, but not until later this year. What has worked is to enshrine in the new legislation our rights to conduct assessment visits, so our power and structural guys can evaluate sites to upgrade for 5G.”

Understandably, there was a lot of talk both at round tables and in corridors about the implications should governments limit use of Huawei products, and if so, how this might impact the timing of 5G rollouts. Huawei equipment is already widely installed in many telecom networks, so any requirement to replace that equipment would be to the detriment of time and money that could be invested into 5G. So timing is one component, but there is also an economic component: the strength of Huawei creates a competitive tension which keeps prices down – if Huawei were taken out of contention, prices could go up and that makes the economics of 5G more challenging.

Government is not just a critical stakeholder in 5G at a Federal / policy level – Municipal governments must be enablers of 5G and smart cities.

“Some cities are investing in communications infrastructure,” said one participant. “For example Nice is making the fibre ducts from their CCTV network available to lower the cost of communications infrastructure. Other French cities are considering the neutral host model; forcing MNOs to share infrastructure. There are transport authorities that own structures and fibre – and of course making that available lowers costs.”

“One of the challenges is that there is no single, consistent and replicable model for making street furniture available to MNOs and neutral hosts,” added another participant. “For example, I believe the concession model has failed in the UK, in part because rolling out the fibre we need to realise 5G is both necessary and disruptive. While boroughs and municipalities want fibre for their citizens, they also want to minimise disruptions – so the reasons beyond 5G to deploy fibre (broadband, small cell deployment) must be leveraged to amplify the case to dig up a road or uproot a sewer – it’s more a civil engineering argument.”

“To many local governments, 5G means risk of more antenna clutter, and increased EMF anxiety. Communications infrastructure becomes politically sensitive – they won’t touch the subject near an election!”

“There are a few success stories of course,” offered another participant. “Sometimes a city starts up a Smart City programme, but if it’s acquired by a commercial company it can make it less open.”

“The communications industry and government share a commitment to create smart cities,” added a panellist. “Capital is important, but every city needs a plan, and few cities have a coherent 5G or smart city strategy. It’s so fragmented. For example, if transport agencies and transport network connectivity programmes were integrated into municipal / borough smart city plans, you could see a very compelling case for 5G, but right now we’re struggling to gain much scale or traction with smart city or small cell trials.”

Neutral hosts can add a lot of value to the 5G rollout if they can bring together the fragmented ecosystem of stakeholders in street furniture – different entities and agencies often control streetlights, CCTV, transport structures et cetera.

Lamp posts can play an important role in 5G. Market leaders Signify (formerly Phillips Lighting) were exhibiting one of their multi-operator smart poles at the TowerXchange Meetup Europe and added “through our lens, communications infrastructure needs to be somewhere. How we did it in 3G and 4G is not tenable – it creates clutter, and we need to protect aesthetics. Whatever infrastructure you come up with for 5G, we’ll hide it in a beautiful streetlight pole! We want to make 5G easier and cheaper to deploy, while overcoming citizen and municipality concerns about clutter.”

“Are there any restrictions on using street furniture for 5G coverage, whether its lamp posts, windmills or factory chimneys – is there a need for a change in legislation to use existing structures?” Asked one participant.

“I’m not aware of any restrictions in UK,” responded a towerco. “EMF restrictions still define a 3-5m exclusion zone, but most of the work to unlock street furniture is not about legislation, it’s about creating an ecosystem of stakeholders and a viable economic model.”

What will be the role of the neutral hosts in putting 5G modules into factory floors? While to date they have tended to focus on landmark hotels, shared workspaces and campuses, we’ve seen pioneers like StrattoOpencell monetising enterprises rather than MNOs to create indoor connectivity – will towercos ultimately expand their focus from purely leasing vertical real estate, to becoming a service orientated business with a footprint indoors as well as outdoors?

“The enterprise business is so different from sales to an MNO,” commented one towerco. “It’s a different sales dynamic, contract tenure is different. Ultimately we pulled out of enterprise because of those differences.”

Other towercos would disagree. For example, the aforementioned StrattoOpencell has been rolled up by Digital Colony, together with more recent acquisitions iWireless Solutions and Spyder, to create a substantial indoor connectivity business in the UK, where towercos Wireless Infrastructure Group and Cellnex also compete for IBS opportunities. That said, the number of neutral host, digital indoor solutions deployed in the UK, still represent perhaps ~500 of the largest premises of a total addressable market perhaps 30-40x that big.

If the jury is still out on whether towercos and infracos should diversify to offer indoor and enterprise services – how about IoT?

“There is a massive opportunity in machine to machine communications: networks designed to support a million connected things in a single square kilometre,” said Oscar Pallarols of Cellnex. “We will each have ten or more connected things at home. If you read and believe any of the IoT forecast reports, there’s still a big chunk of this market to be captured. But it’s not a matter of implementing a new technology; if you put digital pressure on a product it often mutates to a service. Are you ready to provide a service? Do you know who your customer is? Do companies need to change processes to become a service provider?”

An increasing number of towercos are exploring the possibility of locating edge data centres at selected cell sites, exemplified by the aforementioned Digital Colony, by Crown Castle, and in Europe by Cellnex and their Adaptive Edge programme (see “Cellnex: swiftly embracing digital infrastructure innovation”).

“What are edge data centres?” Asked one round table participant.

Roundtable moderator Cara Mascini of EdgeInfra explained; “edge data centres can be located at fibre junctions or along highways where there’s both power and a need for edge compute. Leveraging Cloud RAN or shared macro sites – locating edge data centres near or at the base of towers – could also add value for MNOs and for other companies that need distributed computing. For example, real time gaming can’t be delivered from a core data centre, ideally it should be delivered within 10ms of the end user. Internet giants like Google need to be closer to the end users, and that’s before we even get into the requirements of self-driving cars!”

“The market dynamics are going to be very different in the emerging European market compared to the U.S.,” continued Mascini. “Europe is characterised by smaller, different developments in each country – it’s more homogenous in the U.S.”

One of the challenges for towercos interested in edge computing is a lack of understanding of what an edge computing site might really consist of – and how such sites can be mass produced at low cost. TowerXchange recently published an interview with thought leaders Schneider Electric on this topic, which you can download here.

Another challenge for towercos interested in edge computing are the current differences in valuation between telecom towers, which are characterised by long term contracts with relatively few customers, and data centres, where contracts are shorter, churn is higher, and the customer ecosystems are more fragmented.

A similar factor comes into play when towercos consider diversification into fibre: could it be dilutive to their valuation?

Some towercos believe fiberization of the towers – opening up opportunities to locate small cell nodes along metropolitan and last mile fibre routes – is a complimentary opportunity. Cellnex presumably falls into this category, having acquired Xarxa Oberta de Catalunya (XOC) and it’s 3,000km of fibre for €34mn a year ago. American Tower has also been partnering with, and investing in, fibrecos, although not yet in Europe. Other towercos prefer to focus on leasing towers and rooftops.

“We can’t make investments in fibre work if we’re servicing only MNOs,” said one towerco. “To make fibre work, we would have to go into wholesale and enterprise markets, and that takes us into a different model.”

“We could not get payback in investments in fibre in less than ten years,” complained another towerco.

A fibreco responded: “Our investors are prepared to take ten to 15 year payback cycles – although we expect that to come down over time. Many towercos report demand for Fibre To The Tower (FTTT) but they can’t always make it work because of the return requirement, and because they prefer to invest in their core business. Towercos might be looking for a three to five year return on investment (ROI) cycle, so many are looking to partner with fibrecos that can afford to take longer term view on ROI.”

“If a towerco and fibreco can form a partnership in a sweetspot of densification demand, you have a potential nice marriage,” continued the fibreco. “Is there space in the market for a vertically integrated, towers+fibre play? To do so would be a big task and a big investment – can you execute both equally well?” Digital Colony’s acquisition of Zayo, and their 130,000 route miles of fibre in the Americas and Europe, may prove the concept!

How will demand for fibre be fulfilled? “If you take a dense area like Manhattan the short term drivers are events, new stadia, anything that generates near term demand peaks,” added a fibreco. “We’re working in several major cities in Germany, where the MNOs approach us direct for fibre. They might ask for batches of ~20 sites that they want to refresh: new equipment, update the lease, and fiberize. Around 20 sites at a time is as fast as they typically come, and sometimes that’s sufficient to build a ring so you can densify further.”

Who should pay for FTTT? There was a lots of dancing around that question at the TowerXchange Meetup Europe 2019! “We’ve got to get anchor customers that can support the investment,” said one fibreco. “We’re happy to invest our capital if we have a reasonable view of how to monetise that investment, but we must have confidence of the flow through from end customers. I’m personally comfortable with a mix of customer contracts: sometimes we have a direct fibre relationship with MNOs, who then mix and match their own story of equipment to tower and fiberizing those towers. We see less appetite from European towercos to set themselves up offering a ‘one throat to choke’ turnkey service – although I see some small cell providers keen to step up as a turnkey solution.”

So are towercos prospective partners, clients or competitors of fibrecos? “Through a purely organic sales lens, we don’t see towercos as competitors,” said one fibreco. “Some towercos are becoming more direct competitors in M&A and through M&A, but it depends who buys who. If Cellnex bought a massive footprint of fibre in one of our markets, for example, we’d probably be more nervous! There may be greater ROI on towers than on fibre, but many infrastructure funds want to hedge towers with fibre.”

“The short to medium term drivers for investment in fibre are less about 5G, more about the need for increased consumer bandwidth and for densification,” concluded the fibreco. “You don’t need 5G to need fibre: the need for greater 4G capacity is reason enough to do fibre anyway.”

If the need for greater bandwidth for 4G can drive investment in fibre years ahead of full 5G rollouts, the same applies to edge computing: “Edge compute doesn’t need 5G to have a business case,” suggested EdgeInfra’s Mascini.

“When we took the mobile experience from 3G to 5G, we’ve seen download speeds reduce 20,000x,” said Cellnex’s Oscar Pallarols on one of the panel sessions at the TowerXchange Meetup Europe 2019. “Downloading a movie on 3G took almost a day, in 5G it could take seconds. What we pay for data has been decreasing, and the lifecycle of the asset is getting shorter. How much capacity is still available on 4G networks? We still need to squeeze the maximum from 4G. Average data per user in some parts of Asia is 3x what we see in Europe. That’s why they are deploying so many sites in Asia, as there is no remaining capacity in 4G.”

“The ten members of the European Wireless Infrastructure Association (EWIA), partners of the TowerXchange Meetup Europe, have €60bn of communications infrastructure assets under management,” said Alistair Davidson of Wireless Infrastructure Group, which founded the EWIA. “Initially, in the 2G era, communications infrastructure assets were vertically integrated, and the independent sector didn’t exist. Now independent towercos and infracos are a huge industry in Europe, well supported by infrastructure funds. We stand ready to support with long term, low cost capital; independent infrastructure can lower the economic threshold of 5G by having multiple tenants across one infrastructure; it can produce a higher utilisation, and a higher colocation ratio.”

“We talk about the massive investment needed for 5G, but it goes past passive infrastructure,” reiterated MBNL’s Pat Coxen. “The physical upgrading of macro towers and streetworks, and the swapping of access equipment is a small part of upgrading networks for 5G. We’re anticipating the total re-architecting of existing networks: cloud RAN, software defined networks, virtualised network functions, network slicing, a new DC footprint, a new transport architecture – transmission upgrades will probably be required. That’s just a vanilla list, but there’s lots on it. We’re talking about a full restructuring of the vertically integrated stack. And we need more proof that the capital we’re looking for can realise value: the consumer case for 5G is not yet there. Even the business case is hard.”

“What separates the 5G networks is that they are more end to end contiguous, so we need to be sure we have effective, seamless horizontal and vertical partnerships – from the customer at the front end to crucial enabling systems at the back end. Those kind of horizontal and vertical partnerships are not something happens naturally in most companies,” concluded MBNL’s Coxen.

---

What use cases for 5G are attracting the interest of communications infrastructure owners and investors? Several panellists offered their opinions at the TowerXchange Meetup Europe 2019.

“The simplest use case for 5G is anywhere where you’ve got a high footfall of users,” said Alistair Davidson, Director of Wireless Infrastructure Group (WIG). “When we switched on Anfield, no one told the fans it was switched on, and they drew one terabyte of data in 90 minutes.”

“There may be some interesting 5G use cases such as industrial campuses,” suggested David Crawford of Arqiva. “I see software as an increasingly core part of the end to end service to realise 5G’s very low latency vision. Rolling out 4G was simpler – that was all about coverage for consumers. With 5G there is a consumer use case, but the costs will be huge – fixed wireless access, campuses, industry and factory automation are all potential applications of 5G.”

“I think the early 5G use cases will be in IoT, e-mobile broadband, some smart health, smart city, agritech stuff,” said MBNL’s Pat Coxen. “We see these use cases coming on board quickly, but not ubiquitous. Many use cases will be an extension of what they do already, but with more capacity. It may three to ten, even 15 years before we start to see the truly interesting development of a true 5G economy. Much of the truly interesting stuff is five years out. But we’re building the base building blocks for that amazing future.”

Round table participants challenged the lack of clear use cases to fund 5G. “Gaming is not big enough on its own, Connected Autonomous Vehicles are not yet ready. What is driving bandwidth, specifically paid bandwidth? We’re seeing Internet Giants choke the bandwidth – will they invest in edge data centres?”

“I don’t expect a single use case to drive it all, rather a number of developments that will combine to be the catalyst,” responded another participant. “Gaming, specifically localisation and interactivity and real time-ness, will bring an increased willingness to pay for edge compute and 5G. I anticipate similar developments in the Care market. Automotive is further out but, there’s lots of money there. There’s also private plus enterprise 5G.”

A familiar conclusion came out of all discussion of 5G use cases: collaboration between verticals will be essential to make the whole 5G ecosystem work.

“Cellnex has closed an agreement with a partner managing a racing facility, enabling automotive manufacturers to do the final tests before a commercial CAV is launched to the market,” said Oscar Pallarols. “We’re investing to understand connectivity in autonomous cars. We can think about new sites connected to mobility and how it connects to industry needs. This can help us make decisions related to highways.”

“People want to know what killer app will kick start 5G? Increased capacity, data and speed – it starts with 4G, and we’ve already got customers putting in Massive MIMO, remote radio units, mast head amplifiers, it’s an ongoing evolution,” concluded WIG’s Davidson. “Everything we’re building now we build as 5G ready. Small cells that can support ultra low latency, dense fibre networks that might not be needed now, but the infrastructure is ready to support 5G: the investment is going on now, and we will see it continue to build.”

---