5G: a double-edged problem

How can future networks deliver on the promise of innovation?

I once read that after two years, things rarely look different. After ten years, the shift appears major. This couldn’t be more true of MWC, a yearly milestone for the telecoms industry, which serves as a very public glimpse into the innovations of tomorrow.

5G has been the central topic at the event for some time now, but as we near first deployments, demos and use cases are revealing an increasingly widening gap between innovation and infrastructure.

Roots in Robotics

My own MWC story took root three years ago. To me, MWC17 felt ahead of it’s time: I recall it as the year of 5G, and we were fortunate to be one of the first companies build a demo for an interactive 5G use case.

It was my first experience of the telecoms world — I quite literally found myself thrown into it shortly after concluding a rather riveting Masters in Art Theory — and my god was it fascinating: beamforming? 5G? Killer robots? The focus was on the holy trinity, comprising of URLLC (low latency), mMTC (massive machine type communication) and eMBB (enhanced mobile broadband) as defined in mid 2015, otherwise known as the key to defining the killer 5G use case.

Skip ahead 12 months — in contrast, MWC18 felt all about robots. Robots, robots and more robots, with a handful of drones thrown in for good measure. The need to bring that killer 5G use case to the table felt all the more pressing.

It’s fascinating how much and how little the scene has evolved in three years. Mobile technology is evolving: new hardware equates to higher network demands. For MWC2019, while 5G demos are mocked and become increasingly diluted in messaging (“If I see one more remote robotic surgeon…” — alas, don’t bite the hand that feeds), the hot topic this year was edge computing. The offering? Edge cloud. The word on the tip of everyone’s tongue? Edge. 🔥🔥🔥

5G is (still) coming

We’re starting to surpass showcase moments; moving from the flash of innovation, and crossing into the real-world meaning and experience of that very same innovation. Announcements that captured wider tech press from MWC were not directly operator offerings, but about new gadgets and products reinforcing that 5G deployments will be driven by device adoption ([Magic Leap and SK Telecom Announce Exclusive Partnership | Magic Leap] and [Microsoft announces Hololens 2 at MWC 2019]).

Next, we’ll be seeing workloads move away from these devices: AR is the nearest example of why high density, low latency and compute heavy capabilities will be critical for 5G to deliver on. The success of new devices will depend on there being the right applications that people are eager to use. 5G remains on the horizon, but what is coming with it? The edge is not new, but the set of circumstances that are coming together will make edge computing critical in this increasingly connected world: spatial computing seems to be the premiere-use case for 5G.

Agnostic flexibility is key

The need for agile networks has been written about extensively; the days of proprietary solutions are a thing of the past, next generation mobile networks will be flexible and driven by cloud-native innovations. Open hardware will host a whole range of new software driven services; vendor lock-in is over.

The foundations of the next wave of communication are being built on open, innovative infrastructure that breaks away from the telecoms status quo. Functions that were once slow, complicated and custom now move into the open and are software driven: visibility over the virtualised network offers a better ability to unify and control available resources.

If network function virtualisation is an opportunity for network flexibility, the natural next step for these networks is to ensure high availability beyond basic connectivity. Operators and network owners are uniquely placed to transform the way their customers are served, but whom is 5G for? The immediate benefits are enterprise level, making businesses the key customer (it will take some time for 5G to become a consumer offering with real benefit). Outside of this, there is the entire scope of innovators and service creators left out.

Those who remain underserved

Here we have a very real break from tradition: the two formerly distinct realms of telecommunications and IT/developers are merging from a technical perspective, but what about from a commercial standpoint?

Whilst network innovation is focused on the automation of monolithic, internal processes, there is a gaping hole in the automation of external processes. Third party access to mobile networks is clunky, expensive and for the few, making external deployment an impossibility when compared with the cloud.

What might an alternative process for third party deploying to networks offer? There is an opportunity here to shift who is supplying services to bring mobile operators back into the equation, a shift in both technical and ecosystem dynamics.

Thus emerges the need to define a unified platform that will orchestrate the next generation of communication, that can deploy from anywhere to anything. A platform where mobile operators and network owners can leverage existing infrastructure to monetise third-party access. A platform that unifies edge resources into an abstraction layer, to simplify deployment to distributed locations.

Dynamic scaling

Preparing the edge for third party deployments is easier said than done. Next level services will extend beyond the capabilities the cloud alone offers: applications need localised processing, caching and computing. Backhauling services to the cloud is becoming problematic for dynamic scaling of highly available, localised services.

Looking beyond operational agility, edge hardware and local PoPs alone are not the solution: the non-access edge is an extension of a data centre play, offering only computing resources. It is removing barriers to entry on one level, reducing hardware investment for deployment for those who want access, but what about those who do not want to manage disaggregated computing resources?

At Ori, we believe the distribution of services to the edge is all about multi-access. Preparing the access layer to the network is critical for the edge to become a reality: we need to enable self-service in a digital, on demand world. Offering a new set of tools for developers, we believe deployments need to be more agile, and services need to be orchestrated correctly to extend to the network edge. The beauty? Infrastructure is already there.

Removing the barriers to understanding network layer access is the final frontier for edge computing. There is a very real need for an orchestration layer that allows for fluid access to networks under a unified, edge enablement platform, and that’s exactly what we are building.

For MWC20, we’ll be ever closer to widespread 5G availability, and the tsunami of innovation expected to come with it — the industry needs the proper access for new ecosystems to be in place for that to happen without skipping legacy players.