Network slicing, edge computing, cloud-native development, unified cloud platforms and automation are keys to the integration of cloud and 5G, said Ron Westfall, GlobalData's research director for global telecom technology and software, in the second of a two-part conversation with Telco Transformation.
All of these moving pieces must be carefully coordinated to fulfill the dream of an agile, software-based, high-capacity infrastructure.
Westfall, in part one of the interview, discussed the mutual reliance of these two parallel, deeply related but discreet technical developments. (See GlobalData's Westfall: 5G & Cloud 'Joined at the Hip'.)
Telco Transformation: How does network slicing fit into the big picture of the relationship between cloud and 5G?
Ron Westfall: It's integral. It's basically the sauce of the capabilities that allows the operators to take advantage of cloud computing resources in a more scalable and agile fashion to be able to handle the bandwidth demands of 5G technology and services. What's different about micro-services, as the term inherently suggests, is that we're migrating away from a more monolithic, one size-fits-all-approach to the applications world.
It also is reversing how the industry evolved. It was basically the IT side of the house would say, "Okay, here is the software that we developed and here is the capability. Now adapt to it. We have to plan on a 12-month development cycle, the service can only do X and Y, so we can't promise Z." Now there's potential to leverage the interfaces, the databases, the server logic, to enable more flexible dynamic offerings.
It can be as simple as allowing a customer to change their existing plan. But more importantly, it can be used by the enterprise world to basically not have to sweat changes.
For instance, for special events IT can request a spin out -- again that elastic scaling of the resources -- and that enables the microservice to address that particular event. Another example is if there's an uptick in orders at a company that makes electric vehicles. The company may need a dynamic supply chain to meet this uptick, and if we can do it in a more timely, more competitive way by taking advantage of a microservice architecture built-in to the cloud network then so much the better.
TT: How will the transition to cloud enable 5G? And will 5G help CIO's accelerate the cloud-native movement?
RW: It's important to understand that at the early stages, we had enterprise software, or enterprise applications, that were developed for a non-cloud environment. You have to start somewhere. They're being ported into cloud environments. It's okay when you're talking about limited private cloud implementations. When we're talking about public cloud, and hybrid cloud, and high-scale private cloud implementations, then it does make a significant difference in how the software is developed, how the service is planned, how in essence the operator organization is understanding how to take advantage of the cloud resources to enable 5G.
Cloud-native technology and cloud native capabilities are a strong enabler of 5G. You can still have the debates about how native is this software. But I think that will become less controversial as the cloud universe becomes more prevalent. Basically, you'll be taking it for granted that this software, or this capability, this application, is being developed for the cloud, but with some compatibility for on-premise as needed. It's always going to be based on the customer's need.
TT: What role would distributed edge computing play in regards to enabling 5G applications and services, such as IoT and mobile edge computing?
RW: It's a vital role. We addressed the fact that there are many multiple technological trends here at play to enable 5G. The challenge with edge computing -- just like some of the emerging technologies we talked about, such as NFV -- is standardization.
Quite simply, the edge computing is a part of the puzzle, if you will. It's a technology enabler for being able to quite simply access edge cloud resources in a more distributed fashion. It's the same technique that we're already seeing with caching Internet traffic or video capabilities. It's just doing it on a more massive level. It's on steroids, if you will. Instead of it just being for specific services and applications, it can be done for a software-as-a-service implementation. It can be done for all these different on-demand as-a-service capabilities that the cloud-based 5G universe is ushering in.
TT: How important is it to have a unified cloud infrastructure? And what elements do you think are needed to make that happen?
RW: It's critical to have a unified cloud platform. There will be specific cases where it might not have to happen. But if you're looking at it across the board, certainly on the telco side and most of the customers, it is a key. What's important about unified cloud is that it's enabling the microservices being able to deploy, at scale, individual services on demand. That can be anything from the digital storefront examples or something further out, and vital back-office capabilities such as BSS functions and OSS functions. A unified cloud implementation enables cloud-native implementations that are correlated to being able to auto scale and auto-heal the services. What is different here is that it has to have those automated capabilities, otherwise it just won't work. That is why a unified cloud is important, because it enables the automation techniques.
TT: The close pairing of cloud and 5G seem to be part of the bigger picture of the convergence of telecom and enterprise platforms. Is this true?
RW: Yes. In fact, what we're seeing is there's more blending of these universes. Logically, it's also impacting the enterprise world. What we're seeing now are enterprises on a more regular basis, on a broader basis, using OSS, BSS platforms that were originally designed for telcos.
This is a challenge that is impacting both universes. Quite simply, the cloud can be offered by not just the telco, but by web-scale companies or the enterprise can develop their own cloud or an operator can manage the cloud for the enterprise, and so on. But they're all facing these key challenges on how to scale it, how to secure it, how to take advantage of the agility that's built into cloud computing resources, and being able to enable things like omni-channel journey. Being able to leverage advanced analytics, artificial intelligence to make a marketing campaign smarter, to better personalize a service for a customer or better meet enterprise demands.
— Carl Weinschenk, Contributing Writer, Telco Transformation