By Mika Skarp
I wasn’t in Las Vegas for CES 2017 that closed out this past weekend, but I followed the news closely from home. Although voice UI was everywhere at the show, one of its big winners, Nokia’s Kérastase Hair Coach, an intelligent hairbrush developed with L'Oréal remained silent, and it’s a good thing too. Of all the talking devices out there, people like me don’t need a hairbrush to tell us what we already know.
Silent or not the show’s grand theme trumpeted loudly for the arrival of everything IP. And though there was nothing truly surprising in this year’s show, spare a few utterly ridiculous items like Razer’s three screen, 12 pound, (not so) portable but very stealable gaming console, or a car that attaches to your house to become a room (what if you live on the 5th floor?), the show jumped straight into the waters of IoT with confidence and maturity.
And though the IoT ants come marching four by four, we in the connectivity business can’t help but to keep an even closer watch on the possible cliff of the 5G to be or not to be. And here AT&T delivered, at least in the "need for speed" category.
One of the big differences between 5G and previous generations is that the next G will not be some monolithic roll out of entirely new hardware as it was with GSM, CDMA or LTE. Instead it will be a glorious patchwork, much of it virtualized and software defined. As such, it was no big surprise when AT&T announced that 5G will be here before its own standardization. When, as they proclaim, you are delivering 14Gbps over a wireless connection and less than 3 milliseconds of latency when 5G’s vague standard calls for 5 milliseconds, indeed why wait?
And even though we are now witnessing operator-driven standardization, one thing has not changed and that’s the reality of air interface max capacity. Certainly the race is on when it comes to delivering the highest maximum speed. As cool as it sounds and as easy as it is to market and measure, it doesn’t really matter very much. People (or most people) don’t buy a car based on some theoretical top speed. They put a lot more value on say, an intelligent navigation system that can get you through rush hour as quickly as possible. So while the race for faster over the air speeds seems like an easy sell, it doesn’t address the elephant in the room – that data traffic has grown 250,000% over the last decade and is only going to grow and reliability is what's at stake.
Of course AT&T isn’t only talking speed, but also how these developments will improve connectivity in the home. Case in point; fixed wireless for DIRECT TV. Of course it doesn’t make much sense to sense to bring fiber lines to each and every home where signals will be converted to WiFi in any case. A much better idea is have one eNodeB serving several households using mmWave frequencies. This makes a lot of sense. But in our own experiments we’ve seen that even 4G can be used to deliver an HD TV signal faithfully and reliably leveraging Cloudstreet’s Dynamic Profile Controller ™ or DPC ™ to slice traffic. So as interesting as AT&T’s announcements are, we are already there in the ability to connect residential areas and deliver more than adequate bandwidth to the home combining 5G and this essential traffic profiling feature.
Though they’re generous with details around promised performance, AT&T is not very forthcoming about their mobile 5G trials through 2017. Yes, they’re talking about multi gigabit top speed but what are the applications these dazzlingly fast connections will serve? We can guess that we are talking about super high resolution virtual and augmented reality experiences that run up to 4K and 8K. This is great, but again the issue is not maximum capacity but reliability. Who cares if you can experience a super, duper high-resolution augmented reality environment if you are constantly experiencing drop outs or lag? This will be a non-starter for any of these burgeoning new applications and will have a profoundly negative effect on user acceptance right out of the gate.
The only way to increase reliability for these ultra fast connections is by understanding the capacity requirements of the applications that are using them. Here network slicing comes in to play in a major way. Whatever the speed, it’s high time that we say good bye to best effort once and for all, and when we do the market will surely reward us for it.
Certainly our industry is talking a lot about network slicing, but building an end-to-end capability requires major players like AT&T to support the Proofs of Concept and build the references. And here it is the equipment manufacturers like Ericsson that are defining the standard in three distinct architectural components – Access, Fixed Networks and The Cloud. It will be interesting to see if innovative operators like AT&T bring this essential framework into the fold in 2017 to demonstrate the real power of 5G