The strange dance of 5G

VR in 5G = Vertical Requirements.

This is an edited version of TMN’s weekly newsletter, sent on Friday 17 November. Sign up to receive your copy every week.

Because the press can only write one story at a time, the words of Johan Wibergh, Vodafone CTO, at Huawei’s MBBF last week, were given full coverage.

Yes, he called for a little 5G realism, but really in pretty mild terms. He has in fact done this before, by the way, and I suspect he noticed then that this sort of stuff goes over well. Everyone loves a counter-narrative.

BT’s Gavin Patterson followed up the next day with another slight downer, saying that business models are still required for 5G. But he didn’t say there weren’t any, or that the ones he’d seen were fantastical. “Do we have a business model for 5G?” asked DT CTO Bruno Jacobfeuerborn, “No, but we will have one.”

And yet… BT’s engineers were demonstrating a connection over a 5G Core and a decoupled Uplink/Downlink – that latter part of that demo also mirrored by Vodafone. Vodafone will have most of Milan covered with some element of 5G by the end of 2019, and a fair bit covered by end 2018. And it too has demonstrated UL/DL separation over an “end to end” network (also with Huawei).

Decoupled UL/DL at 3.5 and 1.8GHz

Still, there was a definite sense that the promise of 5G as a vertically-focussed network providing sliceable parameters to self-serving end users is going to be really, really hard to put together.

“We promised a revolution, and we have delivered an evolution.” The words of Antje Williams, head of Deutsche Telekom’s 5G programme.

To precis Williams further:

There were the famous 25 use cases in NGMN’s white paper, and now there’s a triangle of requirements. We promised that everything would be different, and every industry would be satisfied on their different needs. And so far we are all in on consumers.

We cannot, Williams said, satisfy all of our verticals’ needs.

So what are the things that are genuinely going to transform the business, rather than merely evolve it. Three things – network slicing, the network in the cloud, and far higher site density.

These things are revolutionary because they enable the differentiated experiences and vertical industry engagement we have been promising. Cloud brings programmability and flexibility. Density gets the network closer to the user and brings capacity shifts.

NEST at the GSMA

However, the industry has really only made sufficient progress on one of these – Network Slicing. Here the NEST (Network Slicing Taskforce) programme within the GSMA has actually included 10 verticals as part of the group.

But even though Slicing still requires some progress – we need a really big push on the Cloud elements, where cloud-native must be mandatory.

And yet… on the show floor, Huawei’s CloudEngine, its next gen core, its network slicing were all given a run out. The open, cloud-based network has been the focus of the industry for years now.

And we also need to focus in on enabling network density.

Density? “Small cells are very expensive,” said Vodafone’s Santiago Tenorio, citing the costs of leases, permits, infrastructure and transport.

And yet… on the show floor Huawei showed small cells or remote radios installed in lamp poles, with small cabinets at street level, or in the top of repurposed payphone boxes. Vodafone itself is looking at self-backhauling nodes – has been for a whole.

Small cells in a phone booth roof

Radio’s down below, antenna’s in the pole

Using 5G as a backhaul layer for LTE UE Relay

Dense network, Gagnam style.

In Seoul’s Gagnam district, LGU+ is installing 3.5GHz access points every 100 m – 10 of them. In Vancouver Telus has 59 sites in a dense area connected to a pooled baseband, and can achieve 8ms latency for a cloud-delivered service.

China Mobile’s Guangyi Liu, Wireless CTO, said, “Telcos need verticals to break the cycle [of flatlining revs vs capex].” Do verticals need telcos, though? Telcos could offer to share revenues from incremental sales or from efficiency savings made by verticals, Lyiu offered.

Is China Mobile put off by the lack of certainty in this fundamental matter? No. China Mobile is installing 100 test 5G sites next year, and will launch with 10,000 5G sites in 2020.

10,000 5G sites by 2020

Three UK’s Erol Hepsaydir says, “Do verticals know what we can offer them? We need to start talking to verticals, because without getting engaged they will develop something themselves. We need to understand the impact each service experience will have on a vertical.”

Yang Chaobin, head of Huawei’s 5G Product Line, concluded his presentation by saying, “We need to set up a dedicated alliance to define further the requirements of the vertical industries. To feed that into 3GPP. (SOUNDS FAMILIAR)

And yet… In a press briefing held in a side room, Huawei talked of its XLabs project. In a number of labs and test beds it is working with 270+ industry partners on 50 projects to collaborate on different applications of 5G technology.

XLabs is Huawei’s attempt to understand future new applications, creating ecosystems via industry partnerships. By understanding the technical requirement of different types of apps ait can then project those to its R&D teams, especially base station and network R&D teams.

A further goal is to stimulate business models in 5 key areas: connected drones, VR/AR, health, robotics and connected vehicles. Its drone project is called Digital Sky. (Sky Digital, by the way, is something else.)

On stage, on Wednesday morning, in front of a screen as large as a paddock, CEO Ken Hu talked about connected cows. On the exhibition floor you could sit in a drone. You could look at a car being driven remotely from a different location. You could put goggles on and save a football without looking at the actual football itself (something Alan Rough tried to perfect 40 years ago with limited success*).

But a different kind of VR is much harder to demonstrate. And that’s where VR = Vertical Requirements. It is this side of 5G where the industry needs to adopt some strategies, and fast. Or the 5G dance will continue its strange two-step.