The URLLC debate (pt1): Is 5G just a testbed for 6G? (Whisper it, is 5G the new 3G?)
“5G will do the boring stuff first”. The low-latency pyrotechnics that will fire the industrial sector into a new gear for the first time will come much later. The version of 5G that radio engineers are preparing and network marketers are promoting is only a quick-step from 4G. It won’t change anything.
So said the speaker, and the panel at large (featuring BT and Huawei, among others), at the URLCC 2018 event in London some weeks back. It was an essential session, looking at the long-game for 5G – as such, its (rather late) coverage is timeless, at least in the context of network upgrade cycles.
The point of the session was the industry has an intensive workload to actually deliver on the vaulted promise of 5G. The first wave of 5G services will transform neither the industrial landscape, controlling most of the wider economy, nor the telecoms industry itself.
In the end, as this wave crashes, the telecoms market will still be serving subscribers with airtime, as it always has. It will just be dishing-up more for less. The commoditisation of telecoms services, which 5G promises to arrest and subvert, will continue.
The latest set of 3GPP specifications for 5G communications (Release 15, completed June 2018) categorises a ‘full’ 5G system in three parts, covering enhanced mobile broadband (eMBB), massive machine-type communications (mMTC), and ultra-reliable low-latency communications (URLLC). It will be rolled out in that order.
“5G will do the boring stuff first – the eMBB,” explained Mark Gilmour, director of portfolio strategy at telecoms vendor Ciena, at URLLC 2018 in London. “We understand that, because it is what we’ve always done. But to go from that to a URLLC network, which will look vastly different to the eMBB network, will have major implications.”
It might be noted, the middle mMTC phase, was defined in 3GPP Release 13/14, covering low power wide area (LPWA) technologies like LTE for machines (LTE-M) and narrowband IoT (NB-IoT). These are already in play in the 4G era, and bring new use cases.
But the idea of 5G speaks of timeless and limitless communications that will transform society by animating intelligent transport systems, lights-out industrial operations, and smart cities. Most of the sci-fi use cases associated with these can only be served with URLLC networks.
The first 5G terminals will be go-faster smartphones, said Xueli An, principal researcher at Huawei, in London. The production line for URLLC-enabled terminals, ranging from cars to industrial machines, has not been switched on yet. The operator community has to hammer out a plan with industry, starting with the highest-value verticals first.
“We change phones every two-to-three years, and cars every 10-15 years. In manufacturing, machines run up to 30 years. So these machines have very different lifecycles, with very different implementations for 5G modules. Some sectors might not want to upgrade until 6G is ready. These things have to be discussed.”
And there it is: the elephant in the room, the prospect of failure – that 5G, like the UMTS system a generation before, might just be a proving ground for the genuine article. At URLLC 2018, both the panel and audience toyed with the grim theory, not new, that 5G might only be a glimpse of the future, and change nothing by itself.
Here, in a series of posts, we re-play the London dialogue, which also featured UK network operator BT and Germany-based vendor ADVA Optical Networking, considering the gnarly challenges the telecoms market faces to make ultra-reliable low-latency 5G a success.
The session sought to capture the ultimate challenge for 5G by looking at certain key enabling technologies for vertical industries, including multi-access edge computing, network slicing, and latency synchronisation.
Mansoor Hanif, chief technology officer at UK regulator Ofcom and advisory board member at the UK5G Innovation Network, played chair at the London event, and articulated it best, recolouring a question from the audience about scaling URLLC services when so many different industrial use cases are to be served.
“With such a number of moving parts, and such a range of industrial requirements, is 5G just an advanced testbed for basic URLLC use cases? Will it just be nice ideas, and bits and pieces that don’t work, until we get to 6G? Do we have to go through 5G to get to a true URLLC platform?”