As is traditional in the last edition of Wireless Watch for the year, we have been looking back over the developments of the past 12 months, and have selected 12 events which we believe will have a significant impact on the mobile industry, either immediately or in the years to come.
There is one selection for each month, and collectively, this trip down memory lane highlights 12 of the most important broader trends of the year, as summarized here:
Merger of OpenFog and IIC bodes ill for operators’ role in the edge market
Wider theme: Rising importance of edge computing as a corollary of 5G and cloud networks, but rising uncertainty over the role of the operator in the value chain.
This was the first development of the year which showed how telcos risked letting the edge computing opportunity slip through their fingers. Mobile operators had assumed they held pole position in the public edge computing market because of their control of a host of suitable sites for a distributed cloud build-out (central offices, cell sites and so on), as well as high value connectivity. However, many of the most important edge developments of 2019 were focused on the biggest growth area, the enterprise, where operators’ locations and value chains are often poorly aligned with real demand.
OpenFog had already set out a more enterprise-focused vision of the edge cloud, compared to ETSI’s Multi-access Edge Computing (MEC), and by merging with the Industrial Internet Consortium, it went further down the road of a roadmap that was shaped by vertical industry requirements and deployers, leaving only a marginal role for telcos.
Further events later in the year, such as the creation of the Kinetic Edge Alliance (https://rethinkresearch.biz/articles/kinetic-edge-alliance-latest-sign-that-the-edge-is-slipping-through-mnos-hands/), and the increasing investments by AWS and Microsoft Azure (https://rethinkresearch.biz/articles/reinvent-aws-is-now-an-unstoppable-force-in-the-5g-edge-cloud/) confirmed the complexity of the edge value chain, and the fact that operators will have to be proactive in seeking partners and joining unfamiliar ecosystems, if they are to be more than bit-pipes.
Germany stands up to MNOs and aims to earmark spectrum for industrial 5G
Wider theme: The commercial challenges, for MNOs, of supporting specialized connectivity for many different industries with diverse needs, has driven increased interest in private cellular networks. In turn, that has raised the issue of licensed spectrum being allocated specifically for enterprise use, with Germany’s auction being a trailblazer in this respect. Its decision to set aside spectrum for industrial and agricultural use is sure to be emulated elsewhere (though France has gone for a different approach, by mandating network slicing for MNOs’ licences). The European approach is less radical, in support for mobile enterprise and IoT, than that of China or Japan, but it effectively sounds the deathknell for national, exclusive licences being the only way to allocate beachfront spectrum.
The UK, The Netherlands, Japan and Ireland are among the countries offering, or considering, spectrum that is geared to non-MNO bidders, such as enterprises, private network operators or neutral hosts. In the USA, these industrial players are primarily interested in the licensed and unlicensed tiers of the CBRS spectrum in 3.5 GHz, as a way to be able to deploy their own networks without paying huge sums for airwaves.
5G is here, now we must get to ‘true 5G’ – top 10 themes of MWC 2019
Wider theme: From many different angles, this year’s Mobile World Congress celebrated the fact that 5G services were now commercial in a few countries, notably the USA, but mainly focused on the challenges of moving to the next phases of 5G.
The early deployments use conventional network architectures and focus mainly on familiar user bases and use cases, so the impact is really that of ‘4G-plus’. More radical platform and commercial change will only come with ‘true 5G’, which will be transformational in terms of performance, cost and flexibility because it will be based on a cloud-native core, virtualized RAN coupled with edge cloud, and will be able to support many kinds of network behaviors from ultra-low latency to high availability.
The challenges of these migrations were heavily discussed in Barcelona, from a showcase of emerging open network options, from bodies like ORAN Alliance and Telecom Infra Project, to demonstrations of emerging 3GPP technologies like dynamic spectrum sharing (DSS). Rakuten’s mobile deployment, which promises the first end-to-end cloud-based cellular network, was the star of the show, though subsequently, setbacks and delays have emphasized just how difficult it will be for any operator – even one with no legacy, like the Japanese ecommerce giant – to deploy a fully cloud-native network that can perform to the same standard as a vertically integrated, proprietary one.
Lawsuits settled, Intel gone: Apple and Qualcomm locked into unhealthy co-dependency
Wider theme: the bloodbath in the cellular modem business continues, with Qualcomm still dominant, but smartphone makers seeking greater self-sufficiency.
In April, Apple’s feud with Qualcomm came to an enforced end. The iPhone maker had engaged in a series of legal battles with its modem supplier, seeking to challenge its pricing and its whole business model. It had increasingly turned to Intel as an alternative modem vendor, and aimed to use the company’s 5G chips in future iPhones.
But all that came to an end when Intel’s modem chip reportedly did not match up to Apple’s requirements, and it settled its legal differences with Qualcomm and agreed to use its 5G chip. That prompted Intel to exit the market, removing the most serious contender to Qualcomm in the merchant 5G modem space and reinforcing the latter’s dominance.
The drama highlighted the tough nature of the handset modem market, and how hard it is to outdo Qualcomm when it comes to modem performance. The industry is littered with the scalps of those which have tried and failed – Nvidia, Broadcom, Renesas, even Texas Instruments and Freescale once the 4G era started.
But Qualcomm has its own nightmares now. It may have seen off Intel, but the Chinese market is becoming increasingly difficult amid trade wars and competition from MediaTek and Spreadtrum. And the biggest smartphone makers are trying to reduce costs and Qualcomm dependence by making their own chips – Huawei through HiSilicon, and Samsung through its Electronics arm (though it still uses Qualcomm products in some models). Apple has not given up on doing the same – it acquired the team that had been developing Intel’s product, which means it may only be reliant on Qualcomm for one or two 5G iPhone generations.
Further consolidation of power in this key market is likely in 2020, reducing competition both in chips and in the smartphones themselves.
As Huawei’s predicament deepens, the industry as a whole will be the loser
Escalating tensions between the USA and China, related to trade, geopolitics and national security, had a profound effect on the mobile industry in 2019, with US-driven attempts to bar Huawei and ZTE from 5G contracts in allied countries just the most visible development.
Stand-offs between the USA and China have focused on, among other things, 5G leadership. The USA claimed the prize for deploying earliest, but when the three Chinese MNOs switched on their 5G networks, they did so at scale that was unimaginable anywhere else.
The two giant countries have also been trying to reduce dependence on one another’s technologies and patents in key areas like 5G, threatening to split the world into two technical zones, in a kind of hi-tech cold war. In 2018, the USA temporarily barred its companies from supplying components to ZTE, almost sinking the second Chinese vendor before the sanctions were lifted. But in May 2019, it upped the stakes by placing Huawei on its ‘entity list’, which means any US firm must secure a licence in order to be able to trade with the Chinese firm.
Several large names leapt to break ties with Huawei, including Google – prompting the world’s second largest handset maker to resurrect a project to design an alternative to Android. Some were not even American – ARM also distanced itself from Huawei, only to restore full relations a few months later (a blow to Intel, considering Huawei has designed a high end server processor based on the ARM architecture).
As the year neared its close, there were some signs of thaw in Sino-US relations with both sides stepping back from further tariff war escalation, but Huawei and its customers remained in a state of uncertainty. There were several deadline extensions before full imposition of entity list restrictions, but on the other hand, none of the applicants for Department of Commerce licences had been successful.
While US national operators have been barred from using Chinese kit in critical infrastructure, such as mobile networks, since the Obama administration, smaller US MNOs, and operators in US-allied areas like Europe, Australia and Japan, were left in limbo. The US has put pressure on friendly governments to bar Huawei and ZTE, though only Australia complied outright (with more limited support in Japan and New Zealand). However, many countries, and the European Union, are still working on security reviews which could, potentially, still result in bans on specific vendors, or at least more stringent controls on all of them.
That makes it hard for operators to choose their next RAN vendors with confidence that they will not have to swap them out later, and the dilemma is worse in the core, since some governments are leaning towards allowing Huawei in the RAN but not the core network. The uncertainty runs the risk of delaying some 5G investments, while MNOs are very keen to retain a free choice of supplier, rather than see their supply chain effectively reduced to just three for near term macro networks. Samsung is the gainer, as a potential counterweight to Nokia and Ericsson, and in future, so may be the emerging new class of open network suppliers such as Altiostar, JMA, Mavenir, Parallel Wireless and others.
ORAN Alliance claims breakthrough in ending RAN vendor lock-in
Wider theme: The shrinking number of RAN vendors, exacerbated by potential restrictions on Chinese suppliers (see above), has cast a bright spotlight on open solutions which could expand the ecosystem, driving prices down and innovation up. Several important initiatives started to produce real world trials and specifications in 2019, with the Open RAN (ORAN) Alliance capturing the biggest headlines.
The ORAN (Open RAN) Alliance was formed from the merger of the xRAN Forum, founded on code contributed by AT&T, and the China-centric Cloud-RAN Alliance. In June, it said it was working on a new stack reference design that would end vendor lock-ins and allow truly multivendor networks in the 5G era, with all the impact that implies on competition, pricing and innovation. A major breakthrough on the way to that nirvana was released – a multivendor implementation for the X2 interface which connects two LTE base stations.
This is one of a set of essential interfaces between different network elements, which the ORAN Alliance is looking to make fully open and interoperable. In the past, RANs have remained semi-proprietary and locked-in, despite being based on 3GPP standards, because many key interfaces – including X2 between base stations, and CPRI between the radio unit and baseband unit in a disaggregated RAN – were implemented slightly differently by each vendor, making different suppliers’ boxes incompatible with one another.
The Alliance’s work has seemingly become more urgent because of the uncertainty over whether Huawei equipment will be allowed in 5G networks in some countries. Operators which have sourced their 4G RANs from Huawei are faced with a more difficult migration if they have to select a different supplier for 5G, because of the semi-proprietary nature of X2. This has also led to large vendors introducing more flexibility with overlay solutions based on dynamic spectrum sharing (DSS) standards.
With 5G, Hutchison’s Three operators will finally fulfil their disruptive potential
Wider theme: Hutchison’s Three Group is just one example of how the established hierarchy of MNOs could be up-ended by 5G. New cellular entrants like Iliad, Reliance Jio, Rakuten Mobile and Dish; and challenger MNOs like Three; can use a combination of new spectrum, new virtualized architectures, and a more diverse set of business models to seize competitive advantage and market share from the established market leaders.
Hutchison’s Three group of operators in Europe and Asia-Pacific has never quite lived up to its disruptive potential. The Hong Kong telecoms company launched the subsidiaries in six European countries (Austria, Denmark, Ireland, Italy, Sweden and the UK) at the start of the century and of 3G roll-outs. They have had a strong impact as challengers in their various markets, but have remained disadvantaged in terms of spectrum and market share.
That may change with 5G. A more open approach by regulators to infrastructure sharing, spectrum allocation and M&A will allow the operators to intensify tactics which they already use to reduce cost and grow scale. There are three main ways that Three is improving its position ahead of 5G.
- Like other disruptors, such as Iliad’s MNOs in France and Italy, the Three operators like to minimize infrastructure investments by sharing, and ideally to bulk up through acquisition. In Italy, Three is preparing for 5G with both these activities, acquiring Wind Italy to form Wind Tre, the largest Italian MNO (though at the cost of having to divest spectrum to an even more disruptive newcomer, Iliad); and then forming a new network sharing deal with broadband provider Fastweb. The agreement enables a potentially powerful new competitor in an market where the three established MNOs are already dealing with the challenge of Iliad and very high 5G spectrum costs.
- Three has become increasingly known for high impact, disruptive pricing with echoes of the Uncarrier approach of T-Mobile USA. This has been particularly seen in the ultra-competitive UK, where Hutchison was deprived of its wish to buy Telefonica O2. Unlike in Italy, Austria and Ireland, where Three was able to offer enough concessions to get regulatory approval for mergers (with Wind, Orange and O2 Ireland respectively), in the UK it has had to adopt other tactics. These have included very high profile marketing campaigns for Three’s affordable data, content and roaming propositions. It has led the drive to unlimited data in the UK and claims its customers use 3.5 times more data than those on other networks.
- Another important element of the ‘new Three Europe’ is to improve the 5G position, compared to that in 3G and 4G, by investing cost-effective, flexible platforms and in new spectrum. Again the UK is the best example. Here, Three bought UK Broadband for £250m in 2017. That brought a dowry of spectrum in the 3.6 GHz band, neglected and underused in the UK (as elsewhere) until the C-band spectrum from 3.4 GHz to 4.2 GHz became the primary band for first-phase 5G launches. Three UK has reversed its spectrum disadvantage, as well as reducing its need to spend larger sums on 3.5 GHz in the UK auction, and is now the only UK operator to have at least 100 MHz of 5G spectrum. The ability to squeeze more out of every MHz of spectrum will be further boosted by a modern, virtualized network, including its cloud-native 5G core, to be deployed with Nokia.
Will Dish’s $5bn deal with TMO create the 5G disruptor the USA needs?
Wider theme: The restructuring of the US telecoms and media market has been an ongoing saga for a decade. Initially there was a wave of consolidation of smaller players, mainly into AT&T and Verizon; then those two giants have been making media and content acquisitions to extend their business models as they have come under pressure, in the consumer world, from a reinvigorated T-Mobile with its Uncarrier propositions. Meanwhile, the ailing Sprint was acquired by Softbank of Japan from under the nose of Dish Network, only to be now set – if regulators and lawsuits allow – to be acquired again, by TMO.
This is all happening against a backdrop of rising competition in the USA, mainly from the cable operators’ moves into mobile services, and the potential disruptive impact of Dish, which would gain new spectrum and network sites as a condition of the Sprint/TMO merger, should that be successful at last in 2020.
That deal was reached in late July, as TMO looked for concessions to make to the FCC and to antitrust authorities, which have been concerned at the reduction in the number of national MNOs from four to three. The result was a $5bn deal for Dish to take over Sprint’s Boost Mobile brand and customer base, as well as some divested spectrum and other assets, with an MVNO arrangement on favorable terms. That would make it far more cost-effective for Dish to go ahead with its long-promised but uncertain plan to deploy 5G, by adding new resources to its own patchwork of spectrum, and giving it a readymade customer base and roaming/MVNO base. So that would introduce a new national competitor for the remaining three MNOs.
If the current state legal actions against the TMO/Sprint deal fail, the Dish transaction will be activated, and that will end years of ‘will he won’t he’ speculation about whether Dish chairman Charlie Ergen really wants to be a network operator, or whether the diverse spectrum holdings he has acquired over the years – by taking over bankrupt satellite providers and participating in auctions – were primarily tradeable assets.
The TMO deal includes a pledge to the FCC to cover at least 70% of the US population by June 2023, with speeds of at least 35Mbps. Dish has already issued three RFP (requests for proposal) covering various aspects of its planned 5G roll-out, and halted deployment of its NB-IoT network (its only concession, to date, to FCC deadlines to deploy) in order to move to a full 5G project.
Dish could be a truly disruptive force. On a good day, Ergen can outdo even T-Mobile’s outgoing CEO John Legere in bold strategies and eye-catching statements. And the MVNO deal goes beyond the standard agreements usually offered by US carriers. For instance, the companies have agreed to promote embedded SIM (eSIM), which will make it easier for customers to switch operators and port their phone numbers. And it is an ‘infrastructure MVNO’, which means Dish will build out some of its own equipment, giving it greater control of its costs and its quality of service than if it were entirely reliant on a host network.
Over time, Dish can, if the sums add up, shift the balance of its network from MVNO to MNO, as Rakuten is doing in Japan, allowing it to build out gradually and in line with demand and revenue. That will be enabled by its own converged core with which it can provision the eSIMs and subscriber information, and control the balance of traffic between its own cells and those of TMO.
With WiFi 6 and CBRS coming to market together, convergence is in the air
Wider theme: Unlicensed and dynamically shared spectrum is an increasingly important element of the cellular landscape. LTE can be deployed in these bands, and 5G will be from Release 16 onwards. Innovative spectrum sharing schemes are emerging, which provide greater protection from interference than fully unlicensed, because users are prioritized by a spectrum access system (SAS). The most important of these is currently the USA’s CBRS in the 3.5 GHz band, which has licensed and shared tiers and is being closely watched by regulators elsewhere. Its shared portion will support LTE, and later 5G, enabling new deployers and enterprises to deploy cellular networks, as well as providing low cost additional spectrum for MNOs. But at the same time, the main technology for unlicensed spectrum, WiFi, has its own new set of standards, branded WiFi 6, which bring it close to 5G in some respects, providing advances in data rates, device density, latency reduction and other key criteria, especially for enterprise users.
September saw the official start of CBRS services, and the CBRS Alliance was talking about the progress of its OnGo certification scheme, while its older equivalent in the 802.11 world, the WiFi Alliance, was celebrating the start of certification for WiFi 6.
But seeing these technologies as competitors is an old-fashioned view in a world where both can live in shared spectrum and so the divisions are blurring. WiFi is likely to continue to dominate the home and carpeted office markets, but in industrial, critical, highly mobile and wide area environments, it still has disadvantages compared to cellular.
This means there is rising interest in better coexistence, so that operators and enterprises can deploy systems which play to the strengths of both. This year, the
Wireless Broadband Alliance on the WiFi side, and the NGMN Alliance from cellular , have published two white papers on convergence and the commercial opportunities this could enable in key verticals like manufacturing, as well as smart cities, public hotspots, general enterprise and homes.
China’s MNOs are deploying 5G at vast scale, but will it be profitable?
Wider theme: The huge scale of China’s deployment will have implications for the global ecosystem, accelerating availability of low cost 5G devices and helping to establish global norms for deployment processes like migration to 5G Standalone and the 5G core. But it also highlights the dilemma facing many operators which have been pressurized – by competitive or government factors – to deploy 5G early and at scale. This is the doubt over how to make this 5G investment profitable, especially when there is still life left in 4G expansion, and limited opportunity to charge a 5G premium for the new services.
The scale of the Chinese deployment is staggering. 5G services were available on day one of fully commercial service, in parts of 50 cities, including Beijing, Shanghai, Guangzhou and Shenzhen. At launch, the three MNOs had, between them, deployed almost 86,000 5G base stations and there should be more than 130,000 live before the end of this year. This is despite significant amounts of network sharing – in September, China Unicom and China Telecom announced a deal to build and manage a shared 5G RAN, though they will deploy their own cores.
But despite this vast scale, the Chinese operators will struggle to turn millions of cell sites into improved profits. They learned that lesson when they switched on similarly ambitious national 4G networks, but failed to charge a premium for the new services, and were challenged in turning a decent return on their huge deployment costs.
In the most recent quarter, all three MNOs delivered weak results for the first nine months of 2019, and now they are facing the added competitive challenges of new MVNOs and of the introduction of mobile number portability.
In the short term, the 5G launch does not look set to help boost performance. All three operators launched services priced around the same – starting at CNY128 ($18.13) for
30GB of data, and going up to about CNY599 for 300GB. The new 5G plans are now cheaper than 4G on a per-GB basis for the lower-priced tariffs.
A new entrant, China Broadband Network, also gained 3.5 GHz spectrum earlier this year, but is unlikely to start offering services until the second half of 2020.
Vodafone’s 5G RFI – positive for ORAN pact, less so for open supply chain
Wider theme: Open networks were starting to look commercially deployable as the year neared its close, as several large operators sought to accelerate progress and put pressure on traditional vendors. For instance, NTT Docomo of Japan showed off a pre-commercial 5G network in Tokyo – which will be switched on for next year’s’ Olympics – which runs on multivendor equipment supporting the ORAN Alliance interfaces (see June).
Vodafone and Telefonica both issued RFIs (requests for information) last year related to the Telecom Infra Project’s (TIP’s) specifications for open disaggregated RAN, and Vodafone followed this up in November 2019 with a promise of a large-scale RFQ (request for quotes).
It is very valuable to TIP’s progress and credibility that Vodafone, Telefonica, MTN and others are conducting field trials and discussing commercial roll-outs. Such moves increase confidence in an emerging platform while helping to make smaller vendors viable, with the result of widening the supply chain and boosting competition.
Vodafone’s head of network strategy and architecture, Santiago Tenorio, pledged to issue an RFQ (request for quotes)for OpenRAN technology for its whole European footprint (14 countries). “That’s significantly more than 100,000 sites, and all the technologies are to tender — 2G, 3G, 4G and 5G,” said Tenorio. “We’ve invited the incumbent suppliers in Europe of course, but we’ve also invited the open RAN suppliers.”
This was a bombshell announcement in that it relates to large-scale commercial networks, not to localized greenfield trials or roll-outs, often for rural areas, which have been the main proving ground for the various open RAN candidates to date. It opens up the prospect of the shake-up of the architecture, and therefore the supply chain, which operators have talked about for so long as a key enabler of new network economics. It gives hope to alternative equipment vendors that they might be able to bid for tier one contracts, or at least for a share of such deals, in a multivendor environment.
Re:Invent – AWS is now an unstoppable force in the 5G edge cloud
Wider theme: Operators are increasingly acknowledging that they need to partner with the webscale companies, or risk being sidelined by them – and that is not just in the main cloud, where telcos like Verizon have abandoned efforts to build their own public clouds, but also in the edge cloud, which operators once hoped to dominate via their sites and connectivity (see January). At Amazon Web Services’ annual re:Invent conference, a wide-reaching alliance between AWS and Verizon, centered on implementing AWS edge platforms in Verizon locations, showed a potential template for telco/hyperscale symbiotic pacts.
Overall, AWS pushed further into the edge cloud space, and announced strategic partnerships with other telcos like Vodafone and KDDI too, indicating how operators can work with the webscale giant – but implicitly issuing a warning to those which try to be too greedy about the edge and even the 5G value chain.
Verizon and AWS announced a 5G edge solution that aims to marry the strengths of both companies, rather as AT&T is doing in its own cloud partnerships with Microsoft, AWS and others. But of course, the more powerfully AWS extends its cloud leadership to the edge of the network, the more enterprise customers will fear a lock-in – something that some operators hope to exploit by offering a converged, high quality connectivity solution that links to multiple clouds, and allows workloads and data to be located flexibly.
Operators have had a brief window to leverage their in-built advantage over the webscalers – their distributed networks and physical locations – but that is closing rapidly as the biggest opportunities emerge on enterprise premises, rather than public edge sites on cell sites, and as AWS unleashes its Outposts edge offering with full commercial availability.
Telcos still have an opportunity in the edge computing market, thanks to their connectivity, their locations and their knowledge of operating highly distributed sets of assets. But beyond their own classic applications – faster video or augmented reality gaming for their established user bases, delivered from public edge cloud sites – they will not be able to do this alone. They need to find a good balance of power with AWS, Azure and Google, and also, a way to cooperate with one another, in order to match the webscalers’ global presence and ability to serve the needs of multinational customers.