While the ICT industry has been talking about Cloud for years, developments over the past 12 months signal incredible progress in terms of acceptance, adoption and ecosystem development. It also raises questions we all need to be asking going forward, writes PETER JARICH.
Given the immense potential and promise of 5G, it should be no surprise 5G network innovations have dominated any discussion of mobile network transformation over the past few years: 4G-to-5G upgrades; standalone versus non-standalone; massive MIMO support; virtual RAN support. Yet, where 5G is a generational uplift for wireless networks, it’s important to recognise a more significant technology shift which impacts the migration to 5G, along with legacy wireless generations, future 5G evolutions and services that go beyond mobile: Telco Cloud.
While the industry has been talking about this for years, developments over the past 12 months signal incredible progress in terms of acceptance, adoption and ecosystem development. It also raises questions we all need to be asking going forward.
What is it?
It’s somewhat cliche to highlight the term Cloud as meaningless. And, in all fairness, it has been used to mean so many things to so many people (different locations, implementations, use cases) that it can be confusing.
At its core, however, the concept of operator cloud is straightforward, referring to datacentre resources used to support network functions and operations. It might be part of a private cloud deployed by the operator (the traditional implementation). Increasingly, it leverages public cloud resources (more on this later). It could be centralised in a few locations or distributed out closer to the end-user. No matter how it is built, Telco Cloud needs to be integrated into existing network and service operations.
Definitions and implementations aside, there is one other thing we can say about Telco Cloud: it is the most important dynamic in mobile networks today. Bigger than 5G, IoT, AI or concerns about network security. Why?
Telco Cloud represents a shift in how networks are built, leveraging IT industry practices and assets. The result is a foundational layer upon which all those other technologies will innovate, as well as nearly everything in the wireless network, from billing systems and packet core networks, to base stations and the service core. It is akin to the shift from circuit-switched to packet-based technologies in networks and it is why 78 per cent of operators flag IT transformation and the use of IT as an extremely or very important part of their network evolution strategies.
Telco versus public cloud
Several years ago, I was meeting with a wireless network vendor and asked about its support for telecom network functions deployed on the public cloud. The answer (it’s not possible and it is nothing that customers want) felt somewhat short-sighted. Circa 2020, we can see that it was.
The notion of building network functions on the public cloud might have seemed unrealistic in the past given the requirements of carrier grade networking. But as more and more workloads (enterprise and operator) have migrated there, the nature of the public cloud has fundamentally changed. Scale has grown. Toolsets have evolved to support a wide array of applications. New solutions have been introduced, aligning with operator requirements. Think AWS Greengrass and Microsoft Azure IoT, or more recently, AWS Outposts, Wavelength and Local Zones service launches from December 2019.
We can think about these as tools for enabling a convergence (albeit partial) of the telco and public cloud. More importantly, we can clearly see this is the market trajectory. The number of operators AWS announced as partners in tandem with its launches, Verizon; Vodafone Group; KDDI; and SK Telecom, was impressive. This reflects a larger demand and market movement already well underway: nearly 50 per cent of operators claim to be already deploying network functions on the public cloud, with only 3 per cent having no plans.
What comes next?
Operators are now actively moving their network functions to the cloud. In some cases, it’s their own cloud, in others public. Some are centralised, others distributed. This all represents a paradigm shift from how networks were thought about and built just a few years ago.
Make no mistake, we’re still in the early stages of the shift. And, like any movement still in its infancy, this means that there are more questions than answers. A handful will guide the shape of the industry and how it develops as we look back from MWC21.
The next applications. Operations systems, billing systems and the service core are where operators kicked-off their cloud focus: it is only logical given the software-centric focus of these assets. The RAN is an obvious target in the long-term given the capex and opex it represents for operators. But, how far away is long-term? And how long until public cloud is considered suitable for the RAN (hint: 25 per cent already think they’re implementing it)?
We often talk about 5G as a foundational (horizontal) technology which will enable new innovations we haven’t yet defined. The same should hold for telco cloud. As it matures and becomes more accessible, new ways of using it will emerge spanning use cases, functions and benefits not yet cited. What are they? We will learn soon enough.
Best of friends or frenemies. Operators are actively working with the likes of Amazon, Google and Microsoft to move their network functions into the public cloud. At the same time, they’re looking to build their own cloud assets for internal functions and external services, and public cloud players are looking to deploy their own edge services to the same enterprise targets. The result is often referred to as co-opetition, a tense balance between cooperation and competition. How the line between fast friends and fierce enemies plays out for cloud players and operators remains to be seen. It is certain, however, that we should see signs of how this plays out over the next year.
- Peter Jarich is head of GSMA Intelligence.
- Editor’s Noite:The editorial views expressed in this article are solely those of the author and will not necessarily reflect the views of the GSMA, its Members or Associate Members.