Featured Story
After losing Nokia, crisis-hit Intel seeks network assets buyer
Nokia is substituting Arm-based chips for Intel silicon in its latest 5G products amid talk of a possible Ericsson takeover of Intel assets.
At the turn of the millennium, a number of consultants and analysts grew rich by advising telcos that investing billions of dollars in 3G spectrum licenses, and the networks to support an exciting new range of mobile Internet services, would pay off handsomely. The 3G boom never materialized, and the debt-ridden industry spent years knocking its nascent data business into shape.
Today's 4G mobile networks are slick, high-performance vehicles that transport huge volumes of Internet traffic on a daily basis. But the payoff has remained elusive: Data revenues have merely replaced those from the dying voice business, and sales are under constant threat from Internet companies, regulators and plain old-fashioned competition.
Never really held to account for their misplaced optimism, those consultants and analysts have grown more circumspect in the interim. But the industry is once again being swept along on a wave of enthusiasm for next-generation network technology. 5G mobile, which could see initial launches as soon as 2019, has been heralded as the latest revenue-growth opportunity for both operators and the vendors that serve them. Yet investment concerns mean its deployment is likely to be drawn-out and painful.
Figure 1: Good for Sales? Claudia Nemat, the head of technology and innovation for Germany's Deutsche Telekom, thinks 5G could open up a range of new service opportunities.
Ticking along
From a technological standpoint, 5G is ticking along nicely. The 3rd Generation Partnership Project (3GPP) specifications body, which presides over standardization activities in the mobile industry, last week approved a plan to finalize the 5G new radio (NR) specifications by the end of 2017, six months sooner than originally intended. That augurs well for operators that want to launch 5G services in 2019, rather than 2020, by running the 5G NR over their existing 4G networks in a so-called "non-standalone" 5G deployment. (See 3GPP Approves Plans to Fast Track 5G NR.)
In the meantime, a "standalone" variant, including a next-generation core as well as the NR, is due for completion in mid-2018. Some operators have previously complained that fast-tracking the NR specifications could hold up work on standalone 5G, which they consider the more important version. But Gabriel Brown, a principal analyst with the Heavy Reading market research group, thinks it poses little threat to overall standardization efforts. (See 3GPP Likely to Fast Track 5G NR Specs This Week.)
Figure 2: Detail of 5G NR Workplan The 3GPP's 5G NR timetable shows that the non-standalone 5G NR specifications are now set to be completed by the end of 2017, six months ahead of the standalone specs.
Both the non-standalone and standalone technologies will form a part of "5G Phase 1," which in turn will feature in the 3GPP's Release 15 standards update. That will still leave a lot of work to do on "5G Phase 2," which is supposed to be frozen at the end of 2019 for inclusion in Release 16. "You will find more support in that for IoT [the Internet of Things] as well as for mission-critical and ultra-reliable low-latency communications," says Brown. Yet for all the effort that lies ahead, and the squabbling over NR "acceleration," the industry has displayed remarkable togetherness on 5G so far.
Next page: Bandwidth boondoggle?
Bandwidth boondoggle?
Far less cohesive is the rationale for investing in 5G technology. Like previous upgrades, 5G promises a bandwidth boost over its predecessors, and should theoretically be capable of supporting multi-gigabit-speed connections, making it significantly faster than today's 4G networks. What makes it even more valuable, as far as some operators are concerned, is that it will reduce latency, or the delay that occurs when sending signals over data networks. Thanks to a technique called "network slicing," which will help to guarantee latency at certain levels, 5G operators will also be able to run multiple virtual networks over the same physical infrastructure, with each virtual network supporting specific service characteristics.
These different features will unlock a variety of new service opportunities, insist operators. Among the first will be a fixed wireless access (FWA) offering from US telco giant Verizon Communications Inc. (NYSE: VZ), based on standalone Phase 1 technology and using spectrum in the very high 28GHz band. Essentially, this will be 5G as a last-mile, broadband substitute for costlier, fixed-line technologies, and a pre-standards pilot version of it could arrive as soon as this year. From an investment perspective, FWA is perhaps the most defensible 5G "use case," argues Bengt Nordström, the CEO of the Northstream consulting group. But it seems likely to remain a relatively niche affair, having little impact in more developed broadband markets. (See Verizon Will Pilot 5G Fixed Wireless in 2017.)
T-Mobile US Inc. 's chief technology officer, Neville Ray, is one senior executive who, as a major rival to Verizon, is eager to downplay FWA's appeal. "I think it is a use case on the fringe," he said during a presentation at the recent Mobile World Congress trade show. "If that is all we are doing with 5G we should be packing our bags and going home." (See Ericsson's Ekholm Trumpets 5G Role But Still Lacks Plan.)
Figure 3: FWA? You Might as Well Go Home Neville Ray, the chief technology officer of T-Mobile US, cannot resist having a poke at one of his biggest competitors.
For Ray and others, the much bigger and more exciting opportunity is around what the industry jargon terms "enhanced mobile broadband," or eMBB. As the name implies, this would entail using 5G technology to provide much higher-speed connections on mobile devices. A desire to introduce this capability as soon as possible largely explains the interest in speeding up 5G NR development. Used in conjunction with so-called "mid-band" spectrum (between 3.5GHz and 6GHz), non-standalone 5G technology could be ideal for the rapid deployment of mobile broadband services, says Heavy Reading's Brown.
Yet all the evidence indicates that providing higher-speed services will not fuel revenue growth at operators. Between 2011 and 2015, when 4G networks were being rolled out and improved, Europe's biggest mobile network operators all reported declining service revenues in their domestic markets. "When you are building out capacity for normal broadband, the experience is that people are not going to pay much more and that competitive forces in the market will continue to put downward pressure on prices," says Nordström.
Figure 4: Mobile Service Revenues ($B, Using Current Exchange Rates) Source: Operators.
Despite all the talk about the bandwidth demands of forthcoming technologies, including applications based on virtual and augmented reality, there is also doubt that a newfangled network technology is actually needed for broadband purposes. Thanks to recent and forthcoming standards-based modifications, including the LTE-Advanced and LTE-Advanced Pro upgrades, 4G is becoming increasingly muscular, capable even of supporting gigabit-speed connections. While bandwidth-fortifying technologies such as massive MIMO (for multiple input, multiple output, which essentially means adding antennas to receiver and transmitter devices) and beamforming (which concentrates signals on users) are typically seen as ingredients in the 5G story, Spain's Telefónica is investing in them as part of its 4G program, says Enrique Blanco, the operator's chief technology officer. (See 4.5G Sets High Bar for 5G.)
Even operators that are especially enthusiastic about 5G as a mobile broadband technology concede that 4G may suffice. South Korea's SK Telecom (Nasdaq: SKM) is eyeing 5G as a means of handling the growing consumption of multimedia services, says Changsoon Choi, a senior R&D manager with the operator. But he acknowledges that LTE-Advanced and LTE-Advanced Pro could prolong the use of 4G in lower spectrum bands. Others, and especially operators in emerging markets, are outright dismissive of 5G as a mobile broadband option, at least in the short term. "It is clear that throughput is not going to be the use case because 4G is still not fully used," says Yogesh Malik, the chief technology officer of Russia's VimpelCom Ltd. (NYSE: VIP).
Next page: Slicing and dicing
Slicing and dicing
Enhanced mobile broadband is only one of 5G's service-based attractions, however. To assume the technology will be used in the same way as 4G is to miss the point, argue those focusing on 5G's other capabilities. Among these is a much shorter network delay than operators have seen with 4G. That opens up another opportunity in the market for so-called "ultra-reliable and low latency communications." Forthcoming connected and semi-autonomous cars, for instance, would benefit for safety reasons alone from lower-latency connections.
For Bruno Jacobfeuerborn, the chief technology officer of Germany's Deutsche Telekom AG (NYSE: DT), these latency improvements, supported by new architecture, are of paramount importance. "The big data center will disappear and you will get distributed front-end ones," he says. "Suddenly there is a chance to monetize services from over-the-top players." A gaming company, for example, might value lower-latency connections as a way of ensuring Internet gamers all compete on a level playing field, he explains.
{videoembed|730784}
The ability to adapt a 5G network to suit the demands of a particular customer is what really distinguishes it from earlier network technologies, says Telefónica's Blanco. That makes network slicing absolutely critical to the business case. By taking advantage of the software and virtualization technologies finding their way into the next-generation core, operators will essentially be able to offer many different types of service over the same physical network.
Timotheus Höttges, Deutsche Telekom's CEO, is similarly upbeat when discussing network slicing. "Some [users] need low latency and some need only a signal from time to time, while others want high bandwidth for huge packages or a very fast ping rate for gamers," he said during a presentation at Mobile World Congress. "We can tailor to the needs of every specific use case required; we can even tailor price tags to specific requirements." (See DT Plots 5G Across Entire Footprint.)
Figure 5: Tailoring the Network Timotheus Höttges, Deutsche Telekom's CEO, says network slicing will allow 5G operators to meet a diverse range of customer needs.
With network slicing, then, an operator could offer connectivity for smart meters over the same infrastructure used to support ultra-high-definition mobile TV. Yet while this should lead to efficiency improvements, and speed up the introduction of new services, it is unclear why it would bolster the sales potential of the individual "use cases" themselves.
Even if it can, a number of executives have been fretting that regulation on net neutrality, designed to prevent telcos from favoring one Internet service over another, could pose a barrier to the whole concept of network slicing. Such regulation could also make it hard for operators to charge web companies for guaranteed levels of service, as Jacobfeuerborn proposes. (See Net Neutrality Rules Threaten 5G, NFV – Telenor and Telecom Needs New Net Neutrality Story.)
Next page: Downbeat and disconnected
Downbeat and disconnected
It is not only in these areas and as a mobile broadband service that 5G could prove a topline disappointment. Positioned as an enterprise technology, rather than a consumer one, 5G could certainly have more scope for revenue growth, opening doors in markets for machine-based connectivity and IoT. But network revenues in this area are likely to remain small, according to Northstream. Most operators in future will generate only about 1% of sales from IoT business, said the consultancy in a set of predictions about the telecom market published earlier this year. (See 5G Guru Predicts Rollout Disparity.)
To really flourish as IoT players, telcos would need to become more than just connectivity providers, says Nordström. Telco efforts to diversify into value-added services have previously met with little success, however, and 5G technology seems unlikely to help. "The real [IoT] winners will be systems integrators and consulting firms like Capgemini," he says. "They are familiar with the different verticals and can provide strategy advice."
Want to know more about 5G? Check out our dedicated 5G content channel here on
Light Reading.
One problem for telcos formulating a 5G strategy is an apparent disconnect between the technology experts and the marketers. Asked which 5G service opportunities look most enticing, technologists often plead ignorance about the commercial side of the business. But this implies there is no joined-up approach to 5G -- that engineers are developing the technology without knowing how it will be used.
At least one operator has hinted that it does not expect 5G to spur much revenue growth. EE, a UK mobile operator that is today a part of BT Group plc (NYSE: BT; London: BTA), has described its network investment strategy in defensive terms, arguing that sales would decline even more sharply if it did not plow funds into new technologies. "The only way we've been able to increase revenues or even stay flat is to drive more value into the consumer offering -- and that comes from network investment and innovation," an EE spokesperson told Light Reading earlier this year. While the remarks were not specifically about 5G, they cast further doubt over the technology as a sales opportunity. (See EE: New Tech Is Mobile Revenue Savior.)
Next page: A $300 billion bill
A $300 billion bill
Light Reading hopes that 5G can defy these odds and succeed where its predecessors have failed. But the probability of revenue stagnation or decline will make it hard for telcos to fund the nationwide rollout of 5G networks. Operators typically invest between 15% and 20% of their revenues in capital expenditure, says Nordström, and so actual spending is unlikely to rise significantly unless revenues increase. What's more, only a portion of capex will be available for spending on 5G deployment.
Concern about the cost of 5G is palpable. Citing figures in a 2016 report from Barclays, an investment bank, Jacobfeuerborn says the bill for deploying a nationwide 28GHz-based network in the US would be roughly $300 billion. During his presentation at Mobile World Congress, Höttges estimated that covering the whole of Europe with 5G networks might cost anything between €300 billion ($319 billion) and €500 billion ($532 billion). "It is not just about putting new antennas on rooftops but lots of additional investments and so it is important that spectrum is available at reasonable terms," said the Deutsche Telekom CEO in a plea aimed at Europe's regulatory authorities. (See DT CTO: Costs Must Fall or 5G 'Won't Work'.)
Figure 6: Capital Expenditure as a Percentage of Revenues in Last Fiscal Year Source: Operators.
On the spectrum front, at least, there may be cause for some optimism. Amit Nagpal, a spectrum expert with advisory group Aetha Consulting, reckons that spending on 5G frequencies will fall below levels seen during 4G auctions a few years ago, let alone the "crazy" sums that went on 3G licenses. That is partly down to a glut of capacity in higher frequency bands, which are likely to play a critical role in the 5G market. What's more, while operators will continue to value lower frequencies over higher ones, many already hold licenses for sub-GHz bands and could "refarm" those for use with 5G. That could further depress demand for new spectrum. "I don't see any one particular band raising massive amounts of money," says Nagpal. (See 5G Spectrum to Cost Less Than 4G, Says Expert.)
But that could still leave operators facing a huge bill for network rollout. For smaller players, the investment case looks especially troubling, according to Barclays. Operators will have to generate a return on capital employed, or ROCE (a common investment measure), of at least 8% to be able to finance 5G rollout, according to the bank's analysis. While incumbents such as Deutsche Telekom are likely to exceed this figure, Tier 3 mobile network operators are expected to fall well short of it. And the overall industry barely hits the 8% target, according to projections included in the Barclays report.
The actual investment required could vary wildly, however, depending on the nature of the 5G rollout. With most of the cost in the radio access network (RAN) -- between 50% and 70% of the total, according to Jacobfeuerborn -- a deployment based on very high frequencies would be far more expensive than one using lower-band spectrum. That is purely because of physics: Signals travel much further in low spectrum bands than in higher ones, and therefore a network based on the latter would require more cells and more site equipment. In estimating the cost of a nationwide 5G network in the US at $300 billion, Barclays assumes the very high 28GHz band is used everywhere. Bringing sub-6GHz bands into play could reduce the cost significantly.
This has led to some gloom about the business case for networks based on very high spectrum bands, and particularly those using "millimeter wave" frequencies and "small cell" technologies. "Saying 5G implies a small cell deployment is something that would be worrying to me," says Yves Bellego, the director of technical strategy for France's Orange (NYSE: FTE).
Figure 7: Small Cells, Big Concern Orange's Yves Bellego, shown here speaking at a broadband conference last year, is worried about the 5G cost implications of using very high frequency bands.
Nagpal reckons that very high frequencies will only see rollout in hotspots, such as shopping and tourist areas, and that operators will rely on sub-6GHz spectrum elsewhere. Johann Wibergh, the chief technology officer for UK-based Vodafone Group plc (NYSE: VOD), describes sub-6GHz as the "sweet spot" for 5G. (See Orange Also Objects to 5G NR Acceleration and Vodafone CTO 'Worried' About 5G mmWave Hype.)
The necessary trade-off is a loss of high-speed capability as operators move down the spectrum scale, entering bands where airwaves are in short supply. Moreover, even a "mid-band" 3.5GHz-based network is unlikely to map neatly to the "grid" that operators have established during previous rollouts, using spectrum between the 800MHz and 2.6GHz bands. One low-cost option is the 700MHz range, which has already been auctioned off in France and Germany. But this could make 5G seem little better than 4G from a mobile broadband perspective.
Next page: Disrupting the vendors
Disrupting the vendors
Using low-band airwaves might not be the only way to reduce network costs, though. Innovation on the equipment side could also help to improve the 5G economics without undermining 5G performance. Through collaboration with other industry stakeholders, Jacobfeuerborn hopes to be able to "decouple" the RAN control plane, which is responsible for signaling decisions, from the user plane that actually carries traffic. Doing so, he says, would allow the industry to phase out dedicated equipment and make greater use of "standardized" hardware running more sophisticated software. The result should be a much lower equipment bill.
{videoembed|730996}
Orange's Bellego is eyeing similar changes. "Having more features that are software-based [will mean] we can deploy new technologies at the right cost because there is no need for site visits," he says. "That should ease the deployment of 5G."
Such developments seem bound to put traditional vendors under considerable pressure. And industry giants like Sweden's Ericsson AB (Nasdaq: ERIC) and Finland's Nokia Corp. (NYSE: NOK) are already feeling the heat, both reporting sales declines of around 10% last year. "This will change the landscape and some vendors will disappear and some new ones will come into the game," says Jacobfeuerborn. "It is going to be a disruptive moment in time."
Whether or not these new telco demands have some impact, the harsh economic truth means that Ericsson will probably not see the mid-term boost from 5G it craves. That technology, in all likelihood, will be deployed over a much longer period than it took operators to roll out 3G or 4G, slowly infiltrating telecom infrastructure in even the most developed markets. "For a considerable length of time, 5G is going to be an adjunct to a wide area LTE network," says Phil Twist, Nokia's vice president of marketing and communications. For telecom players still dreaming of next-generation nirvana, it may be an unpleasant reality check indeed. (See Don't Count on 5G for a Capex Boost.)
— Iain Morris, , News Editor, Light Reading
You May Also Like