Your browser is not supported. Please update it.

25 June 2015

ITU sets out its 5G agenda as industry risks upgrade exhaustion

Will 5G be a major leap, or a series of incremental steps to consolidate what is already evolving in wireless networks? It’s one of the most fundamental questions on this topic, and the answer is likely to be ‘both’. 5G needs to be a step change in terms of the use cases and applications it can support, enabling radical new business models. But that vision needs to be delivered by harnessing, as far as possible, technologies which are already in use, bringing together LTE-Advanced, virtualization, WiFi, HetNets, high frequency spectrum – and many other elements which are already well understood and even standardized.

A radical business model with incremental technology would be the reverse of 3G and 4G, which saw a drastic and expensive rip-and-replace approach to the networks, heavily driven by the mobile industry – but, in the early stages at least, largely to deliver extensions of existing services. Better quality YouTube viewing on fast networks is merely an extension of 3G data models – billions of dollars are flowing into the infrastructure to enable that, yet the revenues are scarcely higher.

In the 5G era, operators will no longer be able to sustain a hugely expensive upgrade to deliver the same old revenue streams. Instead, they will need a more efficient, software-driven and flexible evolution of the networks, to enable entirely new revenues.

Eric Hoving, CTO of Dutch operator KPN, had it right when he addressed the LTE World Summit in Amsterdam this week. If 5G is incremental in the services it can support, “we are going to fail”, he warned, adding: “Everything we do needs to start with a picture of the future, which often our industry didn’t do. Our industry is extremely incrementally driven, because most of the companies are stock market-quoted, and the next quarter is the most important quarter.”

The new network must focus on what is needed to support the vision – low latency for driverless cars being the most used example. Hoving thinks this can be done using some technologies which are already being rolled out, well ahead of 5G standards, as long as the new business vision is in mind all the time. “I thing 4G is going to be the horsepower for mobile data consumption for a long while. It’s going to be mixed with WiFi, definitely, and ultimately this will grow into 5G, but first we have to define exactly what 5G wil mean,” he said. It also means “cleaning out” the “spaghetti of core networks” and back end systems which most operators have, or “some of our companies will collapse. There is no way you can continue with this IT shebang that we have. We have to first clean out and virtualize.”

Behind all the whizzy demonstrations of brand new technologies, which some vendors and operators are bidding to make part of 5G, Hoving’s words better reflect the actual views of most MNOs when it comes to a next generation of networks – which will be starting to roll out at a time when many will not have 4G fully deployed, and when only a few will be far down the virtualization and SDN track.

So the combination of radical business/services vision with a building-block approach to underlying networks is striking strong chords, and is even reflected in the first official ITU (International Telecommunications Union) pronouncements on 5G.

This time, the ITU acknowledged in its statement of vision for 2020, there will a far more multi-faceted platform to consider, because it will have to accommodate a large number of new use cases. There will be connected ‘things’ of all kinds, virtualization, massively dense networks, ultra-low power applications, and many other elements in play (but many of those are already being developed independently of any 5G core platforms).

The ITU’s secretary-general, Houlin Zhao, said in his statement: “The buzz in the industry on future steps in mobile technology – 5G – has seen a sharp increase, with attention now focused on enabling a seamlessly connected society in the 2020 timeframe and beyond that brings together people along with things, data, applications, transport systems and cities in a smart networked communications environment.”

These are big words, but the ITU always brings a dose of sobriety to a new standards party. Amid the escalating claims about when 5G equipment will be available (now, according to some vendors) and when services will be deployed (2018, even 2017 in ‘pre-standard’ form), the ITU-R Working Party came up with predictable conclusions from its meeting in San Diego, California. The forthcoming standard will be called IMT-2020, reflecting a timeline for finalizing the specifications by the end of the decade.

The official name will do nothing to stop companies labelling every new piece of radio equipment ‘5G’, but it does start to streamline efforts towards something officially and globally unified. Nevertheless, the ITU’s frameworks are always out of step with the work of technical standards bodies and with the real market – and that may be even more the case with 5G, especially if that ends up building significantly on current innovations in areas like virtualization and small cells.

For instance, the ITU frameworks tend to be conservative in their timescales, but over-ambitious in their core requirements – meaning that real world standards-based kit appears earlier, but without hitting the ITU performance levels. The ITU’s timeline is 2020, but like NTT Docomo in 3G and 4G, some Asian operators will deploy pre-standard 5G before the 2020 date, and do almost as much as the ITU to shape how the new networks perform in reality.

In terms of performance targets, IMT-Advanced (4G) was supposed to achieve at least 100Mbps downlink data rates to a mobile end user and 1Gbps to a stationary one. The candidate technologies were actually LTE-Advanced (Release 10 and above) and WiMAX 2, but the former is only now being rolled out, while real world 4G was based on lower-spec’d LTE (and, in the early days, original WiMAX).

In other words, the market, industry alliances, and technical standards bodies like the 3GPP, decide how far vendors and carriers push a new technology to its limits.

So does the ITU have a real role to play? The answer is yes – it does provide a framework which draws in spectrum allocation, operator requirements, regulatory policies and many other factors on a global basis, which helps avoid fragmentation and keep everyone on a single track which is defined without the same level of vendor politics as the technology standards (though of course the ITU has politics of its own).

Nevertheless, defining the framework for IMT-2020 technologies will be an unprecedented test of the ITU’s processes and value to the industry. It is clear that ‘5G’ will encompass a vastly larger range of spectrum bands, use cases, power levels and (probably) connectivity standards than the relatively single-minded standards of the past, which relied on a single air interface geared mainly to increasing bandwidth/data rate.

The body gave few hints about the specific technology attributes it envisages for IMT-2020. Its next meeting, on July 21, will approve the recommendations of the San Diego event, and more detailed work will be presented at the October ITU-R Radiocommunication Assembly.

It will now work with the telecoms industry and other players to specify detailed technical requirements, and the criteria against which to evaluate candidate technologies, probably by year end. It promised to take into account the needs of a wide range of future scenarios and use cases.

Only then will would-be 5G standard technologies be able to compete to be accredited as official IMT-2020 systems. Not that acceptance by the ITU ensures commercial success – for IMT-Advanced, WiMAX2 was selected alongside LTE-Advanced, while IMT-2000 (3G) included the little-deployed TD-CDMA.

In the absence of many hard details as yet, there was plenty of speculation about how the ITU perceives 5G. The organization itself said it had defined overall goals, process and timelines and that “this process is now well underway within ITU, in close collaboration with governments and the global mobile industry”.

Some delegates to the San Diego summit were more forthcoming – according to the Korea Times – citing the country’s Ministry of Science, ICT and Future Planning, a contributor to the ITU-R process – the ITU believes 5G networks must have the capacity to transmit data at over 1000Mbps to over onem IoT devices within one square kilometer.

However, another report emanating from Korea, that the ITU will mandate baseline data speeds of 20Gbps for 5G, proved to be false. Indeed, the ITU itself told Inside5G that, “as of now, the peak data rate of IMT-2020 for enhanced mobile broadband is expected to reach 10Gbps. However, under certain conditions and scenarios, IMT-2020 would support up to 20Gbps peak data rate.”

It will be no surprise that Korea and Japan are prominent in the ITU-R process, and are already engaged in one-upmanship about the importance of their contributions, as well as their plans to be the first in the world to deploy ‘5G’ services, albeit it with pre-standard kit. Of course, 5G may well be a wireless generation that is shaped by China, whose vendors and operators are also engaged in advanced R&D and standards body participation, though making somewhat less public noise than their neighbors.

Another significant influence on the 5G debate is the NGMN (Next Generation Mobile Networks), whose February white paper on the topic is taken very seriously because it is based almost entirely around real operator requirements. Now the Alliance has followed that up by launching a 5G work program to build on the white paper’s guidelines and feed into standardization efforts.

The main work items for the coming months are daunting in scope, considering the group has to coordinate so many operator, vendor and academic members’ inputs. They are to define business principles (“business models, operator capabilities, vertical industry services”); requirements and architecture; spectrum position for WRC-15 and beyond; and IPR position, including stances on standards-essential patents and patent pooling. As well as defining technical requirements for mobile broadband, vertical industries and security, the workgroups will consider end-to-end architecture design principles and management frameworks, while “ensuring that 5G will be architected and underpinned by solid business principles”.

Of course, 5G will not just be about supporting the business models of service providers – whether MNOs, web players or new entrants. It will also be about new opportunities for equipment vendors and chip suppliers, in an industry where they are seeing margins fall and traditional proprietary platforms give way to commoditized and virtualized ones.

The far wider range of service providers engaged in 5G, thanks to the internet of things and vertical market applications, will be the salvation for many hard-pressed suppliers. Business-to-consumer voice and data services have dominated the mobile world, and in that market, MNOs control the delivery and can negotiate the best prices from network vendors. But in the IoT and 5G, B2B will be more important to growth, and there will be many specialist service providers, while networks may be built or controlled by utilities, municipalities or wholesale cloud providers. At the recent Connected Britain conference, some telcos expressed concern that this would enable large OEMs like Cisco to deal directly with vertical market players, shifting the balance of power away from the MNO.

There are also national interests at stake when it comes to the suppliers which will profit from 5G. The one-upmanship in terms of national or regional R&D investment into new technologies is not just about glory, but about sowing the seeds for local companies to be in the vanguard of developing products and generating revenues. Korea, Japan, China, the US and the European Union are all playing this game, and so is Taiwan, which houses so much of the global device manufacturing business. Though it will lose some of its power in this respect to China, it nevertheless sees huge opportunities in the proliferation of low cost embedded devices and gateways, which will lend themselves from the outset to Taiwan’s white label and ODM models.

The island state’s government has set a target of being in the top two for supplying 5G device chips, as part of a broader set of 5G objectives for 2020. Others are that Taiwan-developed IPR should account for at least 4% of the global portfolio of 5G standards-essential patents (SEP), and 50% of that SEP when related to small cells. The state-sponsored Industrial Technology Research Institute has been working with local vendors to develop 5G technologies.

Bodies like the ITU may be seeking to pull the industry debate about 5G into line with the operators’ calls for a realistic roadmap, but some stakeholders don’t think 5G is needed at all (or at least, not for a while). Cambridge may be the heart of the UK’s hi-tech industry, but the annual Cambridge Wireless Debate (this year actually staged in London) asked the question ‘Do we need 5G?’ and received the answer, ‘No’.

Howard Benn, head of standards and industrial affairs at Samsung Electronics Research Institute, spoke for the motion in support of 5G and was challenged by Tony Milbourn, VP of strategy at u-blox, who argued that standardization work was needed for the IoT, but that did not have to be tied into 5G.

Bob Driver, CEO of the Cambridge Wireless organization, commented: “The vote against 5G reflected a strong feeling that 5G and the future of wireless connectivity in general, was so vital to every industry sector, and every citizen, that the development needed to embrace a wider constituency. It was clear that there is a need to consider the wider business models necessary to lead the industry, rather than only focusing on the technologies.”