“Each network slice will address a unique customer need and because customers have different needs depending on the application, many customers may need more than just one slice.”
Mika Skarp, Cloudstreet Founder & CTO
Reflections on an Historic Conference
By Mika Skarp
Cloudstreet’s stand* in Barcelona was busier than ever this year and it's no wonder; we were showing carriers the first and only commercially available network slicing solution, and a path to a new business model based on scores of 5G network slicing use cases.
(*a term my feet now have some reservations about)
The text to follow is drawn from some of the discussions we had and some of the media we’ve been following, and some of what we learned.
While not everyone agrees about its defining standards, there is consensus that something has to be change in 5G if we want different business results than those of 4G. Enrique Blanco, CTO of Telefonica laid it out plain and clear:
["This is not going to be about new devices and antennas. The radio is not the key topic. There is a lot of noise about the radio but we need to make noise about the whole architecture — for me, much more relevant is the issue of network slicing" Enrique Blanco, CTO of Telefonica]
But network slicing seems to mean different things to different people. So where can we find alignment on how get it done?
Despite all the promise, there is cadre of industry operatives arguing that network slicing can’t be done, or is very difficult, not because of the technology, but because of how carriers run their businesses. One telecom operator we spoke to suggested that they could sell a network slice to Mercedes-Benz, and wouldn’t that be great? Sure, and you definitely could, but it won’t make any difference for the end-user because there are at least three completely different use cases for how Mercedes would use that capacity, and each of those would also need a slice.
Here’s what we envision for the Mercedes-Benz network slicing use case;
1. Software Upgrades: This is a classic example of the capacity sensitive traffic class, where over a short period of time you need to download a huge amount data securely and without loosing packets. [See my slightly technical paper on Differentiating Mobile Traffic Classes]
2. Diagnostics & Maintenance: While this is critical to the health and longevity of your 'Benz, it is neither capacity-sensitive (large amount of data), nor delay-sensitive (as a minute or two or ten don’t matter as to when the data is sent or received), and thus can be handled on a best effort basis.
3. Autonomous, Self-Driving: Of course, a car can also be driven by the network, and from its 360 degree proximity sensors to its navigation system, this would constitute an extremely delay-sensitive use case.
But if all this seems still little while off, it is. Remember that the carrier’s core business is still selling SIM cards with data packages attached. It’s no surprise that these standard retail items are commoditized, because they are generic, one-size-fits-all products.
Selling network slices moves the operator’s business model in a dramatically different direction - from being a low-value commodity, to being a high-value solution. Each network slice will address a unique customer need and because customers have different needs depending on the application, many customers may need more than just one slice. Luckily in many of our use cases, including the Mercedes-Benz example, the manufacturers needs are very similar to those of the customer. But there’s no doubt that in order to transition successfully to selling slices, carriers need to take a deep dive into their customer’s businesses and lifestyles and quickly turn those requirements into quality-of-experience-based capacity products and services.
Of course, taking a the plunge into your customers’ needs demands effort, but the return is huge. Like in the 'Benz example, you can immediately see the relative value of various solutions shake out – for example, and perhaps needless to say, all delay-sensitive slices are by their very nature more valuable than best effort, even if the amount of data that is consumed is the same. And this fact opens up a whole new area of discussion and the suggestion that in a sliced 5G network best effort traffic can be free, providing mobile operators a fantastic new customer acquisition strategy.
Another discussion point came from GSMA chairman Sunil Mittal who put in the plain terms his commitment to end roaming as we know it and the bill shocks that come of it. This has been a topic on the from the beginning of 2G era, but it seems finally to have both the necessary technology and regulation coming and a powerful champion in Mittal. Speaking to Business Standard, he points out that “90 per cent of people turn off their data when they travel".
To my mind it’s not only anti-business – as acceptable roaming prices would see travelers opting in, rather than merely hopping from WiFi zone to WiFi zone - but it’s also deeply anti-consumer.
As a traveler to a foreign city and trying navigating it for the first time, your mobile data connection is more important than ever. Sure, using WiFi helps but it is not handy, and you often find yourself doing the “take a photo of someone else’s phone map” thing. (Photo source: Brendan Tully Walsh Marketing Head, Cloudstreet)
And this situation alone provides yet another great use case for network slicing. One of the great things about slicing is that the charging rules per slice can be different. Even in a 4G network you can easily create a slice for roaming customers where their data is charged (at reasonable price!) locally. So in this way the roaming Network Slice can act just like a large WiFi hotspot, covering not just a coffee shop or an airport, but a whole city or even an entire country.
One of the things we also heard at the show a few times, was “easier said than done”. Of course this is true of most things, but at this point and relative to previous pre-virtual Gs, it’s not that much harder. We can do this. We can change the business model and help carriers do better business. And most importantly, we can align connectivity experiences precisely with what a planet full of mobile users are asking for. Call me optimistic, but we can already see it happening, so it’s long since a pipedream.
As always, we'd love to hear your comments, and we'd also love to share them in these pages if you are cool with that. If you have a comment, please reply and say, "yes, please include my comment in your next newsletter.
Thanks and hope to see you all again soon!