Securing 5G’s future means offering great performance, everywhere
You’ve seen the charts. The ones companies like OpenSignal diligently put together to show that even though all the big mobile carriers have long reached 4G maturity, inconsistency in service performance defiantly (and glaringly) remains.
In the U.S. alone, on a 100 point scale, OpenSignal found a ten point gap in mobile video quality between the first and last ranked carriers. Worse, everyone earned a “fair” rating. Consumers will still tolerate bad mobile video experiences, but this kind of performance just won’t cut it in an industry that is screaming from the rooftops that it is ready to support IoT, mobile gaming, connected vehicle, autonomous cars and smart cities.
Let’s break down the two-fold challenge here: 1) mobile operators do not offer consistent quality from one carrier to the next; and 2) if mature networks can’t support basic video delivery, they’re not ready for much more demanding, latency-sensitive applications. Edge clouds are the answer. The best deployment approach? Well, that’s somewhat tricky, as business models need to be weighed against technical capabilities of the key stakeholders involved.
5G making lots of promises, but consistent performance not one of them
What do all of the most talked about emerging mobile services and applications have in common? They require better performance than networks can offer on their own. Maybe it’s low latency - better than 4G’s 60 ms roundtrip on uncongested networks. Or perhaps hefty processing power. To be sure, this isn’t best-effort mobile video - the stakes can often be mission critical. So what happens when IoT devices deployed across a region span cells with widely varying levels of service? Or an autonomous vehicle hands off to a cell that doesn’t offer the same performance as the previous cell? Or when a smart city deployment falls back on a redundant mobile network that doesn’t deliver the same data crunching capabilities to keep the city running smoothly?
5G allegedly solves some of these challenges with network slicing. But not all slices will be created equal. One operator’s idea of the level of performance needed to support mobile gaming might be different from another’s. Mobile networks will need an equalizer to facilitate app and service providers: an edge cloud.
In wireline environments, hyper-scaled data centers get the job done for applications that bring heavy cloud computing demands. Replicating these experiences in mobile will require distributed micro data centers located at the edge of the network. These clouds will host powerful video and AI processors, specialized traffic management capabilities, cybersecurity protection and much more. Putting this functionality as close to the end user as possible ensures services perform the way they’re supposed to. Sharing it across operators ensures services never, ever miss a beat. And that is the key to unlocking this whole new world of opportunity.
Who owns the edge?
Recognizing the need for an edge cloud is one thing. Figuring out who deploys and owns it is where things get tricky. Can mobile operators go it alone? Will they have to work together? Will there be edge “hotels” (B&Bs?) where multiple operators can co-locate? Will a more edge-focused player like EdgeConneX do the heavy lifting (and investing)?
The player(s) that host the bevy of anticipated mobile services stand to take the lion’s share of the accompanying revenue. Of course, there are substantial costs associated with power, equipment, monitoring and security at these sites. A mobile operator that wants to get into this game needs to be able to justify those costs. Another important consideration is whether other mobile operators will be inclined to buy into an edge cloud site offered by one of their competitors. Or if they’d be more inclined to work with a neutral player.
Edge clouds will not be cheap. Operators may well decide a group investment in a third-party edge cloud provider that lets them amortize costs will be the best approach. After all, this isn’t their core competency. Don’t forget as well that a specialized edge cloud provider can also wrap value-added services like remote-monitoring, high-availability and analytics around their offerings.
Getting to the edge is just the beginning
None of this happens without solid business models - something often missing from the hyped discussions that surround exciting new services. Going after the enterprise/campus market may prove to be a prudent approach that will ultimately pave a path to efficient ubiquity. This will also position operators strongly for fixed wireless offerings that will result in yet another dramatic spike in the volume of new services.
There will not be one proven path to profitable edge cloud deployments. In some cases, CDNs may even have a primary role. New application providers will need to see predictable and reliable delivery. Investments will not come unless the business models can be proven first. Just make sure they account for performance consistency, everywhere.
