Never before have events coincided so neatly to demonstrate the gap between mobile operators’ thinking about 5G, and how future networks will really be deployed for disruptive effect. While Europe’s leading MNOs were presenting a backwards-looking "5G Manifesto" to the European Commission, veiling pleas for net neutrality special treatment with promises of 5G build-out, Facebook was announcing an open source approach to the mobile network, OpenCellular – part of a wider trend which could rip apart the network cost base and the operators’ and 3GPP vendors’ cosy world.
Various developments last week highlighted how MNOs are stuck in their traditional thought processes, to potentially fatal effect. This is clear in many critical areas – investment/return on investment (ROI) models; pricing of new services; architecture; and their actual role in the value chain.
So the European operators’ "5G Manifesto" was not a vision of a whole new way of building networks and delivering services. It was basically a negotiating ploy to mitigate net neutrality rules. The central argument was that strict neutrality rules, if applied to MNOs, would limit their ability or motivation to invest in 5G. This reflected several key preoccupations of the MNOs:
- Can they secure return on investment on 5G capacity investments when much of the service revenue goes to over-the-top providers?
- Can they avoid being just bit-pipes for those providers?
- How can they maintain their connection or data fees in the IoT world?
Operators are missing the point of 5G on many levels
All these are the wrong questions when it comes to the next generation of mobile internet services, and they explain why there is deep mistrust of the MNOs in many of the vertical sectors which hope to harness 5G to become always-connected. Senior players in markets like transport, utilities and manufacturing believe that mobile operators remain fixated on mobile broadband data and will not prioritize services which require different network behaviour (high availability, but small amounts of data, for instance), or different methods of charging.
In reality, those three questions are already being answered, by a new set of companies which are coming from the IT and over-the-top worlds into the heart of the mobile network. As the European operators were rehashing their 4G concerns under a 5G banner, Facebook was showing off its OpenCellular open source network concept; Lenovo was unveiling its OpenPlatform for virtualized telco systems; and AWS was acknowledging that cloud services are already commoditised – but not sounding dismayed about that.
These are the companies which will really drive the next generation of mobile broadband and bring it to the unserved masses and the sidelined vertical sectors. Regardless of what the 3GPP decides about air interfaces, these broader architectures, which turn the mobile network into an easily deployable, commoditised IT platform, are the real revolution. So far, very few MNOs are really driving these changes, despite lip service to Cloud-RAN. Even where they are investing in radical architectures, they are stuck in old approaches to ROI on those build-outs, erecting garden walls to keep out unlicensed spectrum operators, cablecos or OTT providers; talking about NB-IoT service fees based on similar models to those for consumers; worrying about neutrality.
Meanwhile, the barbarians are at the gates already. The over-the-top giants have turned into drivers of new infrastructure. Google, Amazon and Facebook know all about building multi-billion dollar platforms very cost-effectively, and using them to enable a host of other providers. They will now turn that expertise to mobile networks, as those evolve into IT service platforms thanks to developments like Mobile Edge Computing (MEC) and network slicing – the real cornerstones of 5G, not a new evolution of OFDM.
AWS’s Brendan Bouffler, head of the Amazon unit’s scientific computing team, put his finger on the button in a recent interview. He said companies like AWS had to accept that their services were commoditised and customer loyalty could not be assumed, but this made the business one of constant innovation in infrastructure efficiency and in service quality. “We’re constantly holding our feet to the fire and forcing ourselves to innovate, that’s how we keep customers,” he said – a motto which few MNOs adopt as they plead for EU protection against the OTT providers which seek to share in the benefits of their networks.
They are fighting desperately against the tide which will carry them to the world of being a utility, in the same way as gas, electricity and transport operators have become – and in many cases, broadband providers. They have resisted the multi-operator technologies which would have made their networks so much more attractive to high value enterprise partners, yet the real 5G potential for the major cellular network operators must lie in a shift from retail to wholesale.
Network slicing is the great 5G disruptor. It creates the ability to dial up virtual, dedicated slices of the network, on-demand, for a particular customer or application, optimized for that usage – and dial them down again when no longer required. This flexible approach could revolutionize the rigid MVNO deals and allow hundreds of service providers and enterprises to use the 5G network, paying just for what they use and making a host of specialized services cost-effective for the first time.
Some MNOs get this – SK Telecom, always in the vanguard of innovation, last week described a software-defined infrastructure (SDTI) project with Ericsson to support slicing (see below). This shows two old guard 3GPP players adapting to the kind of norms Facebook and Google are establishing, but SK Telecom remains in a small minority. Yet if the MNOs do not invest in 5G and then use those networks to enable new services and value chains, others will do it for them, harnessing unlicensed and dynamically shared spectrum, and technologies from the IT and IEEE worlds, to create an alternative 5G world.
The 5G Manifesto
The 5G Manifesto, presented to Digital Economy Commissioner Gunther Oettinger last week by a group of 17 leading operators and vendors, does not raise high hopes that the new-look "5G operator" is going to emerge in Europe.
Ostensibly, the Manifesto is a roadmap to accelerate testing and deployment of 5G in order to get new services to citizens and to put Europe back in the cellular driving seat. The operators – which include Deutsche Telekom, BT, Telecom Italia, Vodafone, Orange, Telefonica and others, as well as the large vendors Ericsson and Nokia – would commit to cooperating on multi-country trials from next year; to mounting user trials as soon as the standards are finalised in 2018; and to deploy a commercial network in at least one city per EU country by 2020.
On the reasonable side, it calls for coordinated action on standards, spectrum, network deployment and ecosystem development across different verticals. But beyond that, it descends into the same kind of lobbying the operators have been doing for a decade.
The European MNOs always tend to get a rush to the head when they try to work collectively, but this Manifesto is particularly wrong-headed.
Three weaknesses in the MNOs’ position
One, it does not reflect the real objectives of most of the operators, whose priority – outside the showcase 5G roll-outs to keep board directors and PR advisers happy – is to continue to squeeze more capacity, revenues and ROI from LTE for as long as possible. Deutsche Telekom itself said this week that it needed at least a decade of LTE evolution and coexistence after 5G appears, echoing a similar statement by 5G trailblazer NTT Docomo of Japan.
Two, the motivations are not as pure as the top line public statements suggest. There are significant strings attached to the operators’ commitments – not just the call for more spectrum and lighter touch regulation, but (the real agenda) demands for concessions on net neutrality. The draft neutrality guidelines are “excessively prescriptive”, warned the MNOs, and would be a barrier to 5G investment and the provision of vertical market services. The operators have legitimate points, especially when it comes to offering mission critical services to enterprises and public agencies, which will require priority.
And three, the thinly veiled bribes in the Manifesto just confirmed that, when thinking about 5G, most MNOs remain preoccupied with their traditional concerns - protecting their exclusive network and spectrum rights, competing in the same old way.
The details of the proposal revolve around the "5G Action Plan" - a timetable for testing and deploying 5G, which would be welcome if it were related to market need rather than net neutrality haggling. The operators say they would start with technology trials, using pre-standard prototypes, by multi-country consortia in 2017; then kick off user trials once the 3GPP standards are finalized in 2018. The aim would be to have live 5G services running in at least one city per EU country by 2020.
There is a list of caveats attached to this roadmap. One, the industry wants the EU's support to ensure appropriate spectrum is available, with harmonized licensing of the 700 MHz, 3.4-3.8 GHz and higher frequency bands above 24 GHz, across the EU by 2020. That would necessitate a unified EU position at the World Radio-communication Conference in 2019.
Two, the operators want financial support for large scale 5G demonstrations, which they say would encourage vertical industries to get involved in trials. Those verticals may feel like poor relations in a 5G process which is so driven by the telecoms vertical itself, but their digital transformation is critical to the EU’s vision of the socio-economic impact of 5G. The Manifesto suggests that existing EU funding mechanisms could be used to provide grants totalling €500m to €1bn, in addition to proposing a new 5G venture capital fund, to support startups developing 5G technologies and services.
Three, the operators think the Action Plan is dependent on the EU taking a light regulatory touch, to make it easier for essential enablers like spectrum, fibre and small cell sites to be made available. The Manifesto echoes similar demands by operators in the past, as well as their trade bodies GSMA and ETNO, to reform the regulatory framework. The general theme is that the EU should keep its regulatory attention mainly on encouraging competition in retail services, while the rules surrounding sites and infrastructure should be relaxed in order to ease and accelerate network deployments, and support co-investment and network sharing agreements among MNOs.
This last is almost inevitable if there is to be a next wave of infrastructure building, before most operators have seen the return on their initial LTE rollouts. But in lobbying for the kind of sharing which will reduce upfront costs, operators are dangerously close to being turkeys voting for Christmas. Network sharing on a large scale is a natural first step to a wholesale-only, utility model, a massive bitpipe role which most MNOs are still resisting.
And in looking for new cost structures for 5G networks, sharing will not be radical enough – in reality, operators will be forced to adopt the commoditized, open source approach espoused by Facebook’s TIP (Telecoms Infrastructure Project), and by Google and Amazon. That will ease their ROI burdens, but will push them into a supply chain which they no longer control.
While the first three caveats are not unreasonable, even though they betray worryingly old-fashioned thinking, the fourth really highlights how the MNOs have been left behind by modern thinking. They are not yet thinking like open internet players, even though internet access and services, and the IoT, are now their primary revenue drivers.
So they are making connections between infrastructure investment and direct monetization, which have long ago been broken in the non-mobile internet world.
NGMN proposes harmony on 5G roadmap
The heavily operator-driven NGMN (Next Generation Mobile Networks) alliance played an important role in crystallizing industry thinking about the requirements for 5G, when it published its influential white paper in early 2015. Now it has outlined its work program for the next few years, a plan which dovetails nicely with a similar announcement from the Small Cell Forum, whose HetNet and virtualized RAN program also addresses the challenges of deploying modern networks immediately in LTE, while having a smooth path to 5G.
The NGMN aims to minimize regional differences in the progress to 5G, to improve economies of scale and reduce fragmentation. In particular, it is focusing on harmonized approaches to developing standard test methodologies and profiles, and to developing an overall 5G architecture.
According to NGMN’s Phillip Deibert, in an interview with The Mobile Network (TMN), there are three main objectives for the NGMN’s new work program:
- End-to-end architecture: The alliance plans to develop an end-to-end architecture paper to submit to the 3GPP. In other words, the standardization process must not just be global and driven by real operator needs, but it must extend beyond the air interface and the RAN, going even further than the 4G SAE (system architecture evolution).
- V2X: There will be a joint NGMN research initiative involving operators, vendors and the automotive industry. The team will work on technology evaluation of current and future options, security, regulatory aspects, spectrum and other items, with most of the work taking place in 2016-2017.
- Trial & Testing Initiative: The objective is to enable global collaboration in testing activities to support an efficient, successful and timely 5G technology and service introduction. Proof points (functional, performance, interoperability) for technology options and testing methodologies will be defined; technical options will be evaluated technologies will be compared and benchmarked against 3GPP and NGMN requirements; and new use cases will be trialled.
US counterparts recently failed, in getting significant special treatment in neutrality, because of the limited capacity of their spectrum and networks (at the same time that they argue to get their hands on the huge capacity of the millimetre wave bands).
Can neutrality live alongside slicing?
The EU and national regulators are still working on exact guidelines and interpretations of Europe’s neutrality rules, even though these are already in place. The draft regulations are “excessively prescriptive”, say the operators, and "could make telcos risk-averse" when it comes to 5G investment.
One of the points they raise seems very pertinent – will neutrality laws allow operators to offer specialised services for verticals, which may require priority treatment in terms of availability and reliability? Driverless cars and smart grids clearly require a different level of QoS from YouTube viewing, so this point is valid.
“We must highlight the danger of restrictive net neutrality rules, in the context of 5G technologies, business applications and beyond,” the telcos wrote. “5G introduces the concept of Network Slicing to accommodate a wide variety of industry verticals’ business models on a common platform, at scale and with services guarantees.”
In fact, many lessons will be learned from how high quality services are being launched in licence-exempt spectrum, where Wi-Fi and LPWA networks are effectively supporting an early version of slicing, and are blurring the lines between open and closed, licensed-spectrum models.
And there is a growing body of opinion that slicing and neutrality, far from being contradictory, actual enable one another. Pure network slicing, based on virtualization and SDN, can place the decision about how the slices are deployed and optimised in the hands of the customer, not the operator.
TelecomTV points to a presentation last year, by Diego Lopez of Telefonica’s R&D team, which showed a layered architecture that allows guaranteed QoS with truly dynamic resource allocation for slices. This would guarantee that every user application could request and receive the class of service it required – low latency for voice, high availability for cars, and so on – without the operator making any decisions about priorities.
That, of course, could raise yet more potential for disruption of the traditional business and revenue models, but MNOs cannot have it both ways – they can embrace new architectures which will enable them to generate new revenues, but they will have to accept that this will be a world where customers and OTT service providers will have the upper hand.
Not just operators, but the Digital Commissioner himself, will need to look to these new ways of delivering services, and ensure that neutrality is safeguarded while not itself becoming anachronistic by failing to support the huge range of services enabled by slicing and virtualization. The Manifesto will be one input to a public consultation which concluded on July 11 and to a series of high level meetings this week focused on neutrality and regulatory framework reform.
In welcoming the Manifesto, Oettinger largely ignored the lobbying aspects and focused on the draft Action Plan, saying: "5G is needed to move one gear up, and envisage a European continent where everything that can be connected will be connected to offer new products and services and improve our quality of life and our competitiveness. I very much welcome the 5G Manifesto and discussions today with the high level industry group. These will help us focus on the key levers to ensure European digital leadership in 5G. I will come forward with a 5G Action Plan in the autumn.”
Copyright © 2016, Wireless Watch
Wireless Watch is published by Rethink Research, a London-based IT publishing and consulting firm. This weekly newsletter delivers in-depth analysis and market research of mobile and wireless for business. Subscription details are here.