Special Editorial Report – July 11, 2017
5G is being touted as the end-all and be-all of wireless networks. It matters little with whom one chats, everybody has a vision for 5G. Autonomous vehicles, remotely piloted vehicles (RPV, commonly called drones), smart cites, homes and infrastructures; eHealth; artificial intelligence; remote control; virtual everything (VX); personal communications; enterprise mobility…the list goes on and on. And it will all be enabled by 5G, “they” say.
That is a lofty vision for two platforms, 5G and the Internet of Everything/Everyone (IoX), that aren’t even anywhere near being fully defined or deployed. But that is the beauty of it. At the nebulous stage of these concepts just about any claim of what they are capable of can be presented.
For example, the other day I got a feed that claims that the IoX is the killer app and 5G will enable that. How? By providing a pervasive, ubiquitous, interwoven fabric of interconnect for everything and everyone. It is supposed to have ultra-low latency, unlimited bandwidth, be super-fast and cover every square inch of the planet.
The year 2020 is supposed to be a pivotal point for both 5G and the IoX. We are 2-1/2 years away from that. The early claims that, in 2020, at least 50 billion devices will be on the IoX is now more like 10 or 20 billion. The IoX will exist, in its true form only when everything and everyone is on it. That is not going to happen by 2020. Truthfully, I really don’t have a clue as to when it will happen. Or when the moment arrives that we now have the IoX.
I recently spoke with a contact I have at California PATH (Institute of Transportation Studies-UC Berkeley). These guys are at the top of the autonomous vehicle food chain. We were talking about true autonomous vehicles and an infrastructure to support it, i.e., vehicles with no user controls or intervention capabilities. His take on it is that this won’t likely occur until the next century. That is over 80 years away. Even if he misses it by half, it is still 40+ years out. Yet, if you listen to some of the diatribe, driver-less cars are just around the corner.
I get some of that same feedback from my contemporaries in the wireless space. Just because 3GPP approved a bump up in the completion of the non-standalone (NSA) implementation of 5G New Radio (NR) from 2018 to sometime in 2017 doesn’t mean everything else is automatically going to scale proportionately. And this is the non-standard. Where is the standard? Now AT&T is saying that it will lay the groundwork for 5G next year.
There is no doubt that some “5G” or 5G-like services will start to roll out in the next couple of years, especially the fixed P2P and P2MP stuff. But, in reality, the “killer” app and visions of 5G and the IoX are at least five to 10, maybe more, years away. A world of lightning fast speed, low single-digit network latency and “unlimited” bandwidth that connects everything and everyone only exists in theory. There are some bits and pieces out there, in bounded applications, but that is about it. And, some things will never need the bleeding edge (do I really care how fast my smart socks tell my smart washer what setting to use to wash them)? And some things that will be under the 5G and IoX umbrella are already here and working today. Other things will function fine with less than 5G metrics, as part of both the IoX and 5G for years to come (industrial M2M for example).
Let’s take a quick detour to other segments. Heck, I know people who are still recording video to a VCR (not many, I have to admit)! And the shiny, new 2017 semi-smart car I just bought still has a CD player right next to the flash drive port. And, they still make tube amplifiers!
So when someone says to me that the IoX, or 5G or autonomous vehicles are the next killer apps, I have to just shake my head. I have been in technology since the Z80 and public safety radio was primarily 150 to 450 MHz. I have seen a lot come and go.
As an engineer, I understand technology development. Yes, Moore’s law has been a staple for 50+ years, but today it isn’t about doubling much of anything anymore. Some even say it is dead, and I now find myself tending to agree.
To wit, one of the places it has face-planted is in semiconductors. There are segments where semis are bumping up against the laws of physics. Take PC microprocessors for example. They have plateaued, or nearly so, at today’s speeds and performance levels. It has taken years to go from 2 GHz. to around 3.5 GHz. And today, the directions of improvements are mostly tangential – parallel processing, chip strapping, multiprocessors and the like. As well, with some exceptions, the software industry is still hung up on single thread implementation.
It is imperative that the 5G and IoX ecosystems do not, similarly, get hung up on single specs and focus on homogeneity, wherever that path leads. Going forward, at least until we lasso quantum physics, or discover ways to push past the limits of physics as we know it, increased speed will depend on efficiency, implementation, agility, smart allocation, optimization, virtualization. But running, flat out, with a focus on Moore’s law is no longer a long-term strategy.
One thing often forgotten, and just as often responsible for progress, is that there is more than just technology involved. An example of that is the latest generation of smartphones. They are just chock full of the latest technology, yet sales are flat, or even slipping.
So just because the one chip of the NR set is coming out a bit sooner doesn’t all of a sudden shift the whole metric. NR is only one, albeit important, element of 5G. As well, aside from hardware and software, there is politics, varying priorities, governmental regulations, spectrum management issues, even human fickleness There is never going to be a definable “aha” moment where yesterday’s 4G, now we have 5G. The same with the IoX.
It is much more reasonable to expect 5G and the IoX to come in stages, and fits and spurts, and to consider much of what is being said about it being “here” as just hype for the time being. We will just become slowly aware that it exists, bit by bit and someday, when most of the criterion is met – so it will have…arrived!