Everyone was agreed that we would have pervasive high capacity networks that would do things like support device to device connectivity on a massive level. This would drive massive quantities of data, requiring a lot of high capacity, low latency networks.
There were some good stabs as to how we would get there. Asha Keddy of Intel reckoned we would have a highly distributed cloud architecture, cloud at the edge, supporting billions of device connections from licensed and unlicensed spectrum. Shayan Sanyal of Bluwan proposed a vision for backhaul which, in the first instance, drowns areas with high capacity mmWave wireless links on a P2MP basis and then takes advantage of new antenna and radio tech that would make coverage smarter and more directed. Keddy said we would see metrics in networks change, from raw bits per second to bps per joule, bps per frequency, bps per area.
“But why?” asked Kevin Fitchard, co-host of the panel and a writer for GigaOm. “What’s it all for?” Hmmm.
This seemed to bring on a crisis in the room. What’s it all for indeed?
Todd Sizer, resident genius at Alcatel-Lucent’s Bell Labs said he didn’t know what the purpose of a 10 Gbps per user network would be, what users would do with that. “We’re not going to carry 55 inch TVs around in our pocket,” he said to chuckles. He was, though, aware that technology has a habit of mugging those who attempt to define the limits of its progress. Apps and use cases would emerge, he said, as they did for 4G. (One thing he did know, the 2020 network will have to enhance the opportunity for operators to make money, he said, otherwise operators wouldn’t be able to invest to build it.)
GSMA Director of Technology, and host of the panel session, Dan Warren attempted to draw some more answers from his panellists, but didn’t get very far. Nobody seemed to know what exactly the use cases would be (although in fact Keddy had earlier outlined just what she thought – vis, a massive number of massively interconnected devices churning out billions of data points).
Perhaps, Warren posited, perhaps what is wrong is that the name this stuff gets given – 5G – is all wrong. (By the way, nobody really knows what 5G is, although there seem to plenty of research grants and dollars being laid down in its name.)
Warren said the session was deliberately not given the 5G moniker, but rather the date 2020 as a marker, and the term Next Access Networks as a concept. Even the date was arbitrary, though, and really stood as a suitably vague marker for the future.
So by the end, we didn’t know what next access networks would be for, how operators would monetise them, when we might see them, and whether or not 5G is a useful or meaningful term. Despite that, we knew a bit better how they might be built. Weird.
And with that, we stepped out into a brighter future, curiously all the better for realising that nobody seems to know what the hell’s going on any more, even when they do.
