Phase Two of EU’s 5G super-project needs to address commercial doubts

As the European Union-backed 5G PPP program enters its second phase of R&D projects, there is a very different mood among operators in relation to 5G deployment, compared to that in 2015, when Phase 1 was launched. Then, real world 5G was still far enough away that MNOs could either be blithely optimistic, or largely ignore the issue. Two years later, a few of their number – though mainly in the US or south-east Asia – are planning deployments or large-scale trials next year. The MNOs are still divided, but they are split over real commercial issues. A few European operators want to be in the first wave of 5G roll-out – before 2022 – but a larger number are questioning the economic basis for doing this.

In the UK, the operators are taking a largely cautious view. Phil Shepphard, director of network strategy at 3UK, told the recent Connected Britain conference that rolling out 5G networks today would be “completely crazy” and would actually bankrupt many MNOs, because of the large number of cells which are expected to be needed. Density is one of the primary advantages cited for a 5G upgrade, with architectures supporting large numbers of small cells to extend capacity and coverage. But currently, cell site acquisition and rental costs, plus the price of backhaul, make huge scaling-up of cell numbers prohibitive.

“Generally speaking, mobile operators are incentivized not to increase site numbers, because every single site increases operational cost and it has to be paid for, which has to come from revenue generated by the customer base,” Sheppard said. “There is a limit to how far operators can go and still be profitable.”

And the excitement about high frequency millimeter wave spectrum is misplaced for that reason – bands such as 28 GHz have very short range and so are well suited to large numbers of small cells, but the site issues mean “it would bankrupt a lot of operators if we did that at the moment, because the operational costs…to do this type of densification would be completely out of the question”, as Sheppard put it.

His comments echo earlier statements about millimeter wave and 5G-class densification, from Bruno Jacobfeuerborn, Deutsche Telekom’s CTO. Earlier this year, he called for greater industry collaboration to reduce the cost of rolling out a 5G radio network, warning that some traditional equipment vendors might not survive the transition from 4G to 5G, so radical a change in the cost base is required.

If that change is not accomplished, through greater sharing, virtualization and new network architectures, then the bill to roll out 5G will be unacceptable for all but a few operators, limiting the scope of deployment. Some 5G changes are already clear – some sites and backhaul will be reused, and software configurable radios plus virtualized architectures could make it easier to insert 5G into an existing structure rather than start from scratch.

But the thorny issue of fiber for dense network backhaul and for Cloud-RAN fronthaul remains, as does the cost of the active antennas which will be needed for multimode networks covering a large number of bands.

Timotheus Höttges, CEO of TMO’s parent Deutsche Telekom, has already warned that the bill for deploying 5G across Europe could be between €300bn ($328bn) and €500bn ($547bn) if there is not a dramatic change in platforms and roll-out processes.
Jacobfeuerborn’s believes the radio will account for between 50% and 70% of the final bill, because of densification – while virtualized basebands and cores support greater scalability, the network will still need huge numbers of physical radios and antennas. So far, this has been the area where there has been least work on driving low cost, commoditized or open source designs, partly because of the engineering complexity of these items.

“If you ask a vendor what a small cell for 5G costs and they say the same as with 4G, it really doesn’t work,” Jacobfeuerborn told a conference earlier this year. “Most of the costs are in the radio access network and we have to get that under control … Core networks represent less than 25% of the cost. Virtualization won’t solve the economical challenge.”

There are some signs of change, in projects like Lime Microsystems’ open source SDR, which is being heavily promoted by Linux major Canonical and trialled by the UK’s EE. And of course, there is Facebook’s Telecom Infra Project (TIP), which is attracting strong operator support, especially in Europe, for its low cost designs for small cells, Massive MIMO antenna arrays, and other key elements.

DT of Germany, BT of the UK, and Orange are among the backers of TIP (see separate item below), as they seek to diversify their RAN supply chain, encourage innovation in this very closed market, and drive radical price reductions among their established suppliers.

“We have to collaborate as an industry to drive the cost down. We can’t work in the same way as previously,” Jacobfeuerborn told LightReading. “This will change the landscape and some vendors will disappear and some new ones will come into the game. It is going to be a disruptive moment in time.”

Orange’s recently launched Telecom Track initiative will act as an incubator for disruptive RAN start-ups, and connect them into TIP and its significant ability to influence large MNOs worldwide (see Wireless Watch June 19 2017).

It will be important that TIP, or other similar initiatives which are likely to emerge as the RAN landscape shifts, drive unified platforms. Fragmentation is one of the biggest risks in the early days of a new standard, because it reduces the economies of scale that any operator can achieve. A few frontrunners are prepared to take on the cost of customized, individualized systems to enable early moves, but the majority will continue to enhance 4G rather than take on the cost and risk of an immature solution.

This has been the concern of some operators about the recent disputes over how to accelerate first wave 5G standards. Verizon has gone its own way, working with its suppliers to define a pre-standard platform which it can deploy for fixed wireless, in millimeter wave spectrum, as early as late 2017. It insists that it will be able to tweak these systems to align with full standars when those are published.

Rival AT&T wishes to do the same thing, but took a different approach – pushing (successfully, and supported by a band of other MNOs) for 3GPP to fast-track a subset of 5G New Radio (NR) standards so that deployments can start from 2018, but based on official specs. That means that the specs for the non-standalone strand of 5G NR (which still requires an LTE anchor network) will be finished by the end of this year and expected to be approved by March 2018, allowing for standards-based products and deployments in early 2019 or even late 2018 (especially where MNOs want to support fixed wireless first, since this is not reliant on smartphone availability).

Opinion is divided over the merits of each of these fast-track approaches – or indeed, whether there is any real need to move to 5G so quickly anyway. European MNOs have been particularly negative about the arms race which is building up, even if this will be more short-lived than 4G’s WiMAX versus LTE, or the GSM versus CDMA, debates, since all the parties are committed to convergence as soon as possible.

Enrique Blanco, the influential group CTO of Telefonica, told Mobile Europe in May that speeding up the 5G standardization process risked stripping out many of its benefits, and resulting in a platform with limited differentiation from 4G.

“My personal view is those operators who are really pushing for this – in the US, and Japan, and Korea – have different, more urgent priorities,” he said. “Because of the Olympics, and whatever else, they want to deploy services using NR much sooner – which means they need the standards to be defined sooner, and the chipset makers to deliver sooner.” But he added: “5G is not just about sub-millisecond latency and 1Gbps speeds – we can get close to that with 4G already. So what’s new? The business model needs to be defined; 5G shouldn’t just bring new radios and antenna systems but new network architecture.”

Orange’s director of technical strategy, Yves Bellego, also thinks focusing on a limited range of radio specs carries high risk. He said in an interview: “5G is not and should not be limited to NR. We really believe we need a full 5G system… and there is a risk that defining a new radio and not addressing other portions does not in the end deliver the best solution. Having new radio does not answer all the needs.”

Even temporary splits over standards can encourage fragmentation, and delay the attainment of economies of scale. Howard Watson, CEO of UK telco BT’s technology, services and operations division, used his presentation at the recent 5G World event in London to call on chip providers to focus on global standards, not country-specific deployments.

He wants key chip developers to prioritize global spectrum bands, whether millimeter wave or sub-6 GHz, rather than chasing one-off projects like Verizon’s (though he did not specify that deployment). He said focusing on individual operator platforms would delay availability of affordable “global components”.

“We need to set a baseline with 4G and build on that to transition to 5G,” said Watson.
“Next generation networks will be based on customer needs, and not be technology-led.”

The main drivers of globally uniform platforms and components will be 3GPP and -given that 5G will be a whole architecture and not just a radio – standards bodies and open source initiatives with their roots in the Internet or data center worlds. But there is also an important role for R&D programs to set a common agenda. Work by Europe’s 5G PPP and its equivalents in the US, China, South Korea, Japan, Brazil and other key regions has been pre-commercial but highly influential in setting priorities for vendors and standards groups – especially given the high level of cooperation between different regions’ initiatives to achieve some global consensus.

Indeed, the 5G PPP is possibly the biggest 5G effort in the world and has ties to many others round the world. There are co-development agreements with Brazil, Korea, Japan and China, among others, and talks to include India and the US in the family.

In particular, these projects and cooperations are important for defining the broader architectures around the core radio standards.

Now these initiatives tend to be entering a more commercially focused phase as standards come closer. Phase Two of 5G PPP is a good example. The European Commission has committed €700m to the new stage of this public-private partnership (PPP), which is the 5G infrastructure strand of the €80bn Horizon 2020 scheme. There will be 21 projects, following the 19 which were funded under Phase One. The biggest shift in focus in the second round is towards software-driven and cloud-based networks for 5G.

Phase Two, like its predecessor, is designed to develop innovative technologies and promote new use cases. It was kicked off last week during the EuCNC event in Finland, a timely launch given the recent tide of criticism, from many operators, that the EU’s policy makers were moving too slowly on 5G and ceding commercial advantage to other regions.

These complaints were largely related to spectrum and regulatory policy, beyond the scope of the 5G PPP, but additional clarity on architectures and economic cases will certainly help regulators with their deliberations on a harmonized structure across the EU.

Even Phase One was not all about bluesky research, but included a large number of feasibility studies, including a detailed economic analysis by Real Wireless. Under the auspices of the 5G PPP, there will be a 5G showcase at the Euro 2020 football championships, which will take place across 13 countries. And the European Commission set out a 5G Action Plan last autumn, which will leverage the 5G PPP research. Its agenda is to:

Align roadmaps and priorities for a coordinated 5G deployment across all EU Member states, targeting early network introduction by 2018, and moving towards commercial large scale introduction by the end of 2020 at the latest.

Make provisional spectrum bands available for 5G ahead of the 2019 World Radio Communication Conference (WRC-19), to be complemented by additional bands as quickly as possible, and work towards a recommended approach for the authorization of the specific 5G spectrum bands above 6GHz.

Promote early deployment in major urban areas and along major transport paths.

Promote pan-European multi-stakeholder trials as catalysts to turn technological innovation into full business solutions.

Facilitate the implementation of an industry-led venture fund in support of 5G-based innovation.

Unite leading actors in working towards the promotion of global standards.

And Phase Two of 5G PPP will add to the store of knowledge and understanding, with most of its projects due to be completed in the second half of 2019, with five continuing to 2020. Eight of the Phase One projects – most of which will be completed by the end of this year – will evolve into Phase Two projects, while Phase Two will also have 13 new items, chosen from 101 submissions.

According to the EC, Phase Two has several key focuses:
Leverage Phase One results
Shift emphasis more heavily towards software and cloud networks in the RAN and core
Extend inclusion of vertical industries which are expected to be heavy 5G users, especially automotive, healthcare, energy and media
Extend involvement to new stakeholders and domains (such as satellite and optical)
Represent small and medium enterprises (SMEs), which are expected to do over 20% of the work
Drive Europe towards 5G trials and demos

The project leaders are clear that they will adhere to 3GPP and ITU IMT-2020 standards and do not intend to do anything pre-standard, in a bid to ensure unity across Europe. This will please the European MNOs which criticized the Verizon approach and the 3GPP fast-track, but others may feel they want to move more quickly with showcase, pre-standard deployments and that the EC is not helping them achieve that.

Perhaps the most important of the Phase Two goals to make 5G roll-out commercially worthwhile is the extended involvement of vertical industries. Although most 5G projects pay lip service to the need to support the requirements of many sectors, not just telecoms, in reality there has been frustration among organizations which represent industries such as transport and manufacturing, which feel their priorities are placed below those of MNOs.

It is also notable that over-the-top and cloud providers are not specifically mentioned in the list of objectives. Although some of these companies do participate in 5G PPP projects, the member lists for most of the initiatives are packed with traditional MNOs and their established vendors. Meanwhile, Google and co are increasingly driving the traffic on the mobile networks, and monetizing it, while Facebook is spearheading a whole new supply chain with TIP.

On the vertical industries front, one of the Phase Two projects is 5G-Transformer project, which aims to develop a new NFV/SDN-based 5G mobile transport and computing platform to support flexible network slicing, especially for low latency use cases, and to enable the federation of resources across multiple domains. The 30-month project, which started this month, has a grant of €7.98m from the European Commission. The consortium comprises 18 organizations, including NEC, Ericsson, Nokia, InterDigital, Orange and Telefonica.

The concept will be designed, developed and demonstrated for three industries – health, automotive, and media & entertainment, as well as for telecoms itself in the form of a flexible MVNO platform. The project will use established frameworks – NFV, SDN and Multi-access Edge Computing (MEC) – to create a platform to create and manage transport slices. The team envisages three main architectural building blocks – a vertical slicer; a service orchestrator for end-to-end service orchestration and federation of resources from multiple domains; and a mobile transport and computing platform (MTP) to control underlying and integrated fronthaul and backhaul.

The project will build on work done by the Phase One 5GEx project on cross-domain orchestration and federation mechanisms and apply them to the mobile transport network. In addition, it will extend the design created by the 5G-Crosshaul transport Phase One item, by integrating a MEC service platform to support vertical industries.

Such projects can easily be dismissed as too slow, too academic, too bluesky, but they do bring competing vendors and operators together in a neutral environment and help enable important testbeds and other prerequisites of commercial confidence in new platforms. But the MNOs know that the real threat is that those over-the-top suppliers, or vertical industries, will feel sidelined from the mobile industry’s 5G activities and take the initiative themselves, coming up with money-making services – perhaps harnessing shared spectrum – before the MNOs do so.

Some MNOs are certainly keen to outpace the EC program and take their place in the vanguard alongside the Japanese, Korean and north American pacemakers. Traditionally, the Scandinavian and Finnish operators have been early movers, helped by the proximity and investments of Ericsson and Nokia. TeliaSonera was the first MNO in the world to have large-scale 4G commercial services, just pipping Verizon to the post, and Telia was prominent in the campaign for a fast-tracked 5G NR Non-Standalone spec last year. The operator says it will launch services in 2018 based on these specs (as far as they are finalized at the time), initially targeting the cities of Stockholm, Sweden, and Tallinn, Estonia.

And some regulators are taking an aggressive view on enabling early 5G deployment. In France, regulator ARCEP said last week that 5G services could be ready to launch in France during 2018 (see separate item).

But despite all the activities of regulators, 5G PPP, the EC Action Plan and individual operators, it will be hard to match the momentum which is building up in China, where the government and the operators are determined to be world leaders, and to reverse the technical and IPR reliance they have had, in previous generations, on foreign organizations.

Chinese players are amassing large stores of 5G-related patents, contributing to many standards and 5G industry initiatives, and also developing advanced business models.

Last week, the research division of China’s Ministry of Industry and Information Technology (MIIT) released a white paper that predicts 5G technology in China will drive CNY6.3 trillion ($930bn) of economic output by 2030. By the same date, 5G-related sectors will generate around CNY2.9 trillion of economic value added and 8m new jobs.

The biggest beneficiary, the report concludes, will be equipment and handset makers, serving the domestic and export markets. The second biggest beneficiary group will be Internet companies providing 5G-enabled services, from ecommerce to content streaming to B2B software.

The third group will be telcos, notably placed behind the over-the-top and cloud service providers. However, analysts at Jefferies believe the government will provide considerable incentives to the operators to ensure they build out their 5G networks quickly, and identifies China Unicom as the main recipient. In a client note, analysts Edison Lee and Timothy Chau wrote:“If the government wants Unicom to build a 5G network of reasonable scale so as to compete with CM (China Mobile) and CT (China Telecom), we believe lowering its cost structure by giving it the spectrum at 700 MHz is the only way to achieve that. The entire industry’s capex will likely not fall a lot but all the savings could go to Unicom, which should be a boost to its valuation.”

China’s MIIT recently issued a public opinion request for 5G spectrum at 3.3-3.6 GHz and 4.8-5 GHz and followed that up with a request for detailed feedback on the potential use of higher frequencies for 5G, namely 24.75-27.5 GHz and 37-42 GHz. These fall within ITU global frameworks, with the higher frequencies due to be discussed at the next World Radio Conference in 2019.

In December, China issued new Radio Regulations, which shift away from the traditional reliance on administrative approval and adopt a more market-based approach involving auctions. This could enable a new entrant to the market, and relate spectrum fees and holdings more directly to business case.

The changes are part of a broader move to lay the groundwork for China to be a very advanced 5G country, both in services and equipment provision. The country’s IMT-2020 (5G) Promotion Group announced technical specifications in November for its second wave of 5G tests, with technology R&D tests taking place between now and 2018, and product R&D tests to follow in 2018-2020, in time for commercial launches in that latter year.

China Mobile says it will test “experimental” 5G base stations in four or five cities next year, in sub-6 GHz spectrum. China has been promoting a domestic 3.5GHz specification for 5G networks, which will also be adopted by Softbank in Japan and possibly Sprint in the US’s CBRS band. By 2020, China Mobile aims to roll out up to 10,000 5G base stations.

“Having learned the lesson from promoting its own standard in 3G (TD-SCDMA), China is now highly determined to support global compatibility in 5G,” the Jefferies analysts wrote. “This will support China’s ambition to become a leader in the 5G global supply chain.”