Since the dawn of the internet era, there were networks and applications that ran on them. The two were separate and distinct. Virtualization and the cloud are erasing that dividing line, creating incredible possibilities for service providers and businesses in all sectors, while also opening the door for disruption by allowing non-traditional players to get into the “network” business. If there’s anywhere these changes are going to play out most vividly, it’s in the emerging edge cloud.
Virtualization in the service provider world really took off toward the end of 2018 with a dramatic jump in software-based, “cloudified” network functions. Since then, network services have essentially evolved into discrete consumable items like many other cloud applications. This puts cloud providers in a position to offer network-type services and means traditional service providers need to be strategically prepared to compete.
At the same time, mission-critical business applications have become increasingly integrated with and dependent on network services, meaning they now have performance and reliability requirements similar to traditional telco offerings.
Business applications with network-like demands and network services that are consumed like applications are at the heart of concepts like 5G network slicing: service providers will be able to dedicate on-demand “slices” of their networks to specific customers with tailored latency, throughput and more. To deliver those performance characteristics, some network service components will have to run in the edge cloud, as close to users as possible.
Taking the network to the edge
As network functions move to the cloud, their location can be optimized; it’s no longer necessary to route all customer traffic through centralized data centres for service processing. This is taking the edge cloud, where distributed network and service applications are hosted closer to the end user traffic, from proof of concept to reality.
While the edge cloud’s “killer app” is still to be determined, service providers and businesses are already exploring its potential with a number of use cases where low latency wireless communications are a must, including connected vehicles, automated manufacturing, streaming media, cloud gaming and more.
Decentralizing and distributing cloud resources will add complexity to the network, especially from a service design perspective. At Nokia, we’re investing in ways to mitigate that complexity. We’re driving forward with end-to-end automated orchestration, not just for service creation but also service delivery, assurance and monetization. We’re also moving our own portfolio to the cloud, building network functions that let businesses seamlessly transport workloads from any cloud to any cloud as needed and, eventually, have workloads automatically move themselves to the most appropriate clouds for their requirements.
The time to prepare is now
The 5G-enabled edge cloud will support a vast range of new, low-latency services and applications. But there are still a lot of unknowns. How close will the edge be to the customer’s premises? How much investment will be required? Will cloud providers like Amazon and Microsoft partner with or disrupt traditional network providers?
What does seem clear is that the edge cloud isn’t likely to be one single “kind” of cloud. A hybrid private-public cloud solution can take advantage of the dynamic, on-demand nature of the public cloud to manage cyclical peaks in demand while providing sufficient redundancy for resilience. This will minimize the issue of over-deployment of resources within a single cloud environment.
With network-like apps and app-like networks running in distributed clouds right to the edge (wherever the edge ends up being), there’s no telling what innovative service providers and businesses might do. The time to prepare for those possibilities is now and to start imagining new experiences rooted in ubiquitous network resources and propel the edge cloud into reality.