BY MIKA SKARP
5G is getting hyped in all kinds of interesting and very futuristic ways. Every day seems to arrive with a new set of never-thought-of-before scenarios. But it got me thinking; how many of these use cases are actually related to 5G, and how many of them are really about applications, transport and the cloud?
Looking at the Internet on its own, we see three layers at the physical level; access, transport and the cloud. While with 5G we are mostly concerned with changes to the access layer, and particularly via the air interface. But of course this isn’t the only game in town when it comes to access. Remember that only 25% of all Internet traffic is generated by macro access networks, also known as “the Gs”. So that leaves 75% of traffic, give or take, that is generated solely by WiFi. And yet, there’s not that much ballyhoo about new WiFi standards like 802.11b ac etc despite how cool they are.
But of course, that ratio will shift in the direction of macros access networks with the arrival of more capacity. That will in turn set the requirements for the transport layer as it connects access networks to the cloud. At the moment, commercial deliveries of single Lambda Switches capable of carrying up to 100 Gbps are in play with precious little fanfare. Certainly, the same goes for the cloud, where capacity is increasing by leaps and bounds. Of course, the biggest investments in that regard are happening in the new data centers of Google and AWS, but that isn’t public information and for very good reason.
Operating Systems, programming languages, databases and all manner of different frameworks are developing in scores of open source projects around the world. And yet, while these quickly evolving systems will have an enormous impact on, say, autonomous vehicles, and will be mandatory to develop new applications, they are seldom spoken of. So again, why so much hype and hoopla around 5G and something that represents but a fairly small piece of puzzle?
Perhaps the most obvious reason is that 5G has simply been oversimplified. And in so doing it has made a loud but perhaps dubious case that it will make all the difference. At the moment 5G is defined in most discussions as a new air interface between the handset and the base station. But as we accumulate rather than replace Gs, it does make a lot of sense to talk about them; and it certainly is fun especially in the abstract. Much of the dialogue seem to exist at an almost philosophical level where many mobile pundits find their happy place and where many a policy maker feels like they 'get it'; after the 4th Generation will come the 5th, and after that the 6th. No doubt but that this is a good marketing strategy and certainly does full duty in keeping these technologies in the headlines. It also allows us to talk about technology development without going in to the terrible technical details that put policy makers and the public to sleep. Rather, it allows us all to have a conversation that stays nicely bubbly and afloat at the conceptual level.
But there’s one big fly in the ointment; somebody needs to actually turn these concepts and abstractions into a reality. But the industry is changing, and that's opening up completely new technical avenues to companies that not only understand the complexities, but can also deliver solutions at pace. Certainly, the biggest of these opportunities comes of the fact that we are no longer talking about the physical layer of our networks as we have in the past. As any manufacturer knows, developing a new RF chipset, processor or laser switch is cripplingly expensive. But as we move into the realm of the virtual and the all-important NFV and SDN frameworks, a little clever network virtual architecting later and you’ve turned the whole industry upside down.
We can think about connectivity as a sort of fossil fuel for the information age. While our networks grow faster than Oil is being depleted, that’s not where the comparison lies. It’s rather more about how we can extract ever more from something that is limited by nature. If we imagine the Gs to be like oil drilling equipment, only instead of sucking gallons from the ground it’s about pulling Gbps out of the air. But we need to remember that consumers care as much about that as they do the latest groundbreaking alternative energy initiatives. They care about getting from a to b as quickly and cheaply as they can, and that goes for cars out on the highway as much as it does devices out on the mobile Internet.
Certainly, the next decade will bring us as many new discoveries in an “all of the above” strategy for digital transformation. And with it, we’ll see a lot of consolidation, (horizontal and vertical) and likely even mergers between OEMs and Cloud providers. And most certainly, yes, there will be a 6th G, and in a fully virtualized macro network stack, it may just come a lot quicker than any G before it.