news


Telefónica explains why it won’t be rushed into launching 5G

5G phone IoT cloud

Amid the rush to shoe-horn 5G into every telecoms announcement, Spanish operator Telefónica thinks the industry should take its time over this crucial technology.

In an interview with Telecoms.com, Telefónica’s Ignacio Berberana, Innovation Manager, Radio Access Networks Division within the Global CTO team, shares his views on the pace of 5G development, standardization and expectations.

Telefónica’s research and development arm, Telefónica I+D, represents the company in global and regional fora dedicated to advancing technology in the telecoms industry. Technically, Berberana falls within I+D, but his role is mainly within the group-wide CTO team, where he is the director of all things related to the mammoth radio access network operation the telco runs around the world.

Today, Telefónica is dedicating several projects towards 5G both internally and on an industry-level. Telefónica I+D is the entity participating with 5GPPP projects, and Berbarana says they are very involved in a few different areas within the organisation right now. On top of that the telco has a range of internal innovation programmes to define 5G and test some of its components.

There’s also the recently opened 5TONIC open lab in Madrid, which Telefónica is planning to use in collaboration with different members of the ecosystem, not just vendors but also companies in vertical markets, industry and local governments.

It is also involved with 5GIC at Surrey University, but because of the uncertainty surrounding O2 in the UK the powers that be also felt the need to stack its chips on a project in its home market of Spain.

With such a long road ahead and wide array of technologies which need development before they can be called 5G; where does one of the world’s largest telcos, which operates in such a broad array of global markets, begin?

“Firstly, we think that 5G will start in low frequency bands, so our primary concern from a radio viewpoint is in the support of technologies that may help us to increase spectral efficiency in low frequency bands, but not to the extent that we compromise backwards compatibility with LTE,” says Berberana. “In low frequency bands we are looking for the support of massive IoT services, which we fully expect to benefit from 5G. For massive IoT it’s not so much a question of capacity of spectral efficiency, it’s more about signalling efficiency. With billions of devices connecting we don’t want the network to be useless, so the area of connection-less support is something we’re focussing on.

“The other main area is using Massive MIMO to increase capacity,” he says. “We have been involved in some activities with some of the European projects and trials with some vendors and we’ve seen some promising results so far.

“Following that, we think that there needs to be a new way of supporting mobility, in the sense that it should not be linked to a specific cell, more to the user that is roaming between connections. We call this the cell-less radio access network, and we think this is a good idea if we are going to have a network with Massive MIMO support where the cell edge is not so clear.”

This explanation from Berberana suggests Telefónica is not so much adopting a revolutionary approach to its technology research, but instead a progressive evolution of technologies which would be compatible with existing infrastructure. This strategy appears to be at least in part driven by the current absence of 5G standards definition, so it looks like Telefónica is looking at its own needs right now, and developing use-cases based on feasibility of being included in future standardisation frameworks.

Berberana says the work Telefónica is doing on Massive MIMO and spectral efficiency is something it will benefit from in the long term regardless of whether or not standards bodies adopt them as pillars of 5G.

“Obviously we are not in the position of some other operators who are saying they will be supporting or rolling out 5G services pre-standards,” he says. “We are supporters of the standards, but of course we only want the best technology components to be selected and this is an area in which we want to heavily cooperate with vendors and forums like the NGMN. “

2018 has been a widely touted target for some operators looking to get on 5G fast. SK Telecom has already committed to a live network deployment, at least in trial mode, in time for the PyeongChang 2018 Winter Olympic Games, while MTS in Russia has said it wants to have some sort of 5G service ready and in stadia in time for the FIFA Football World Cup in 2018 as well.

“Having Phase 1 by 2018 and Phase 2 by 2019 is quite ambitious, and we are not quite happy with that,” he says. “We would be happy for a more relaxed framework for defining these standards, but it’s been agreed and we will support that.”

With Berberana’s hesitation towards rushing a 5G deployment for the sake of getting it done quickly, he does not appear too bothered about rushing to hitting the 5G milestone as early as possible.

“I guess it may be feasible to have pre-standardised systems in place for those dates,” he says. “It may happen, but we are not looking at it like that. In the initial conversations we’re having with customers in industry or government, the famous goal of 1 millisecond latency is useless for them; on the other hand they do want deterministic latency. By that, I mean us promising 10 millisecond latency and the customer getting 10 millisecond latency. Not 5ms latency half the time and not 15ms the other half, it will be 10ms.

“This is a strong requirement, but the problem we see with this rush to 5G is that it might not cover all of the features it needs to cover for all of the users.”

The narrative that’s beginning to form behind Berberana’s words is one of reliability. Speed and capacity and latency all feature as an important facet of 5G, but faultless reliability appears to be the underlying theme.

“90% of mobile users might say they want improved speed, and improved capacity; but if you want to go to a manufacturer in industry you have to provide something different. Deploying some of the functionality we’re seeing demanded of by 5G will be very expensive, and it will not be feasible to do so in the short term.”

Speaking of speed, the marketing hype surrounding 5G is promising unprecedented mobile download speeds, but it has repeatedly been tempered by talk of the fifth generation of mobile technology being about far more than just a big pipe, speeds’n’feeds job. That being said, European and Asian operators have already begun trialling or, in some cases, deploying LTE networks which they claim deliver download speeds greater than 1Gbps to the newest handsets in optimal conditions.

Berberana has a somewhat more modest expectation, taking a relaxed stance over the potential download speeds the everyman will experience on a 5G network.

“There is talk of gigabit speeds for enhanced mobile broadband experiences, but we think that in at least the first phase of deployment between 2020 and 2025, these sorts of speeds are unrealistic,” he says. “This would require a transport network which can handle this sort of traffic, which is very expensive to deploy. You also need a lot of spectrum available to support this for a huge number of users, and it is unlikely regulators will allow us to use it.

“Right now, you have network speeds of 30-40Mbps, and we would be happy if, by 2020, we could get 300-400Mbps. Perhaps in very good conditions you could get gigabit speeds.”

The concept of having a variety of “very good conditions” is something Berberana wants to see come to an end. By that, more specifically, he wants to see everyone get the same, equal speed wherever they are in relation to a cell.

“The thing we want to see, and this is something we also said when we were defining 4G, is equalising the received performance for users and that still hasn’t been achieved. The user very close to the base station gets very good performance, while the user further away gets an order of magnitude lower capacity. Getting a continuous experience wherever you are is far more important than getting the fastest speed.”

If operators are going to be processing a, presumably, exponentially growing amount of data; and if Berberana’s comparatively modest expectation of 300-400 Mbps for every user is even vaguely feasible for users on the street; it is head-spinning to consider the backhaul challenges in play.

“Backhaul is the most complicated area, for sure,” he says. “Here you have additional requirements; you not only need a high capability network, you also need centralised processing which can coordinate the cells providing coverage in different areas. In that sense it is very, very complicated and we expect to use high-frequency bands above 100 GHz which may allow you to implement unilateral, low cost, high density networks to support this. Obviously the main solution for this is to have a high capability fibre network.

“Some people are also promoting 5G as an alternative to fixed, and that can be problematic because you need the high capability fibre network to support the sort of mobile speeds and coverage we’re talking about.”

With talk of highly efficient fixed networks being so integral to the success of 5G, thoughts turn to the concept of 5G acting as a replacement to traditional fixed-line home broadband once the successor to LTE becomes mature. Statistics released by the US government shows American households are increasingly turning to mobile as their primary means of internet connectivity, could that trend continue in a 5G era?

“Not yet, to be honest,” says Berberana. “The kind of technologies we are developing for 5G may help bring connectivity into the home in a cheaper way. Deploying fibre inside the building is quite expensive and challenging, not only in terms of work but also other non-technical considerations. So in that sense we think that beamforming capabilities for high frequency bands may help reuse this technology for other purposes. But as far as the idea of having 5G for domestic services goes, we don’t think the complete end-to-end solution can be based on mobile networks for the time being.”

So where does Berberana think Telefónica will deploy 5G first, given it is active in so many markets?

“We have operations in developed and undeveloped markets, and it has always been the case that we have deployed new technology in the mature markets first,” he says. “For specific 5G use-cases, it’s not just about increasing the capacity or spectral efficiency, it’s also about providing wireless services at a lower cost to incorporate low-income customers and providing coverage to rural areas where it’s been traditionally difficult to do so.

“In summary, we’ll be looking at our mature markets to start off with, Spain, the UK and Germany, as well as some areas in LATAM, such as Brazil, Chile or Peru. That all depends on the availability of spectrum we need to do so, which is difficult because of local regulation and the nuances between them. If we had the opportunity, that’s where we want to start.”

Perhaps even more so than with previous generations, the telecoms industry is investing a lot in 5G and is understandably keen to promote its next big thing. However, as Berberana explains so well, a lot of things need to be done properly and coordinated globally for 5G to be what everyone wants it to be. This will take time and rushing it out before it’s ready is likely to be an own-goal.

  • 2020 Vision Executive Summit

  • LTE Advanced Pro and Gigabit LTE: The Path to 5G

  • TechXLR8

  • 5G North America

  • 5G World


One comment

  1. Dave Burstein 07/06/2016 @ 8:07 pm

    This is a very strong interview with important, little reported data. Thank you.

Leave a comment

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Polls

How have open source groups influenced the development of virtualization in telecoms?

Loading ... Loading ...