Throughout the ‘ask-the-experts’ session at URLLC 2018 in London earlier this month, serialised here across a number of posts, the conversation circled back time and again to 5G use cases. What are the sync requirements? How many slices? Where is the edge? Every time, it seems, the answer depends on the use case.
It presents a chicken-and-egg investigation, with many of the use cases still emerging. Can the telecoms industry wait around for the use cases to present themselves? It is the opposite of the old build-it-and-they-will-come movie line.
This impasse also begs a definitive question about the future scalability of ultra-reliable low-latency communications (URLLC) services – that ultimate strand of 5G that promises to transform whole industrial sectors with machine intelligence, and society at large through such thing as driverless cars and intelligent cities.
Ofcom technology chief Mansoor Hanif, an excellent chair of the London session, picked the question from the floor, posed in response to the revolving conversation, and put it back to the experts.
“How do we get that scalability, like with previous generations, when there are so many different use cases, and they are all so very different? Could it be that one day, we just close the door, and say, ‘from this day on, these cases cannot be monetised?’”
Milan Lalovic, principal researcher for 5G mobile core research at BT, returned to his point about opening up the edge network for developers to spark new innovation. “Again, this is why it’s so important, for example, for us to look at ETSI MEC APIs because there are challenges there,” he said.
The operator community must work out the gnarly question of discoverability and distance between edge nodes, he said, so applications can auto-select appropriate compute resources.
“All these challenges, we are hoping to work through the industry through some sort of common APIs, which we can provide to application developers to define those use cases,” he explained.
Hanif rejoined, to make clear the message. “So having standardised common APIs is key to getting that scalability?”
“Yes,” said Lalovic.
Again, Hanif summed up. “It’s a good question because the whole point is to open up to the use cases and sources of revenue. But you need a single platform to get the cost efficiencies, and you need to be flexible enough at the same time to handle the particularities,” he said.
The debate around URLLC is how to achieve a mix between a standardised platform for everything, and a flexible platform for everything.
This fascinating conversation wound inwards again, and circled back to its start, about whether 5G in the end will unfold as a proving ground for 6G, which will deliver finally on the hyped promise of industrial transformation. It is a contentious question, said the audience member.
“I have a view on that,” said Anthony Magee, principal engineer at ADVA Optical Networking, also on the panel.
“I don’t know if necessarily we will be talking about a complete rip-and-replace when it goes from 5G to 6G. I see it as much more incremental. So, whether the marketing teams get hold of it and call it 6G in the future is a whole different argument.”
“They will,” said Hanif.
“But I think the industry has been more sensible this time around,” said Mark Gilmour, director of portfolio strategy at Ciena.
“We are not mandating ripping out all of the LTE infrastructure that is in place. So I think your question is not that contentious. It just points to the fact it will be a journey over the next few years.”
Magee zoomed out, and would it up. “If you look at the 5G story, this is the first time people have looked at the wider transport network, and not just the radio interface. I look at 4G and 5G, and I could put in edge compute and improved synchronisation, and I could enable quite a lot of URLLC-like features on the transport domain,” he explained.
“I don’t need a 5G air interface to do that. So for that reason, I think the transformation of the network will continue apace, and whatever happens on the radio network will happen as well. The two will just continue along – 6G, 7G, whatever else.”
Is 5G just an advanced testbed for 6G? (pt1)
Why is ultra low-latency 5G so hard? (pt2)
How will 5G slices be provisioned? (pt3)
What is the business case for MEC? (pt4)
Will enterprise 5G use cases scale? (pt5)
Table of Contents
