In 2019, as virtualized network architectures are rapidly adopted to support 5G, we expect to see containers emerge as the de-facto platform to run new applications and workloads.

The excitement around 5G is building as we hear more news about network deployments, trials and handsets. However, one 5G-related issue that hasn’t yet been crystallized is what form 5G software and innovations will take, and how these new services and applications will be deployed into the network. Unlike 4G/LTE network infrastructure, the architectures that support 5G are virtualized and cloud-based, so the smart money is on application developers, mobile operators and equipment vendors using microservices—and, in particular, containers—to drive 5G evolution.

It makes sense to use containers to support 5G, as they will provide operators with a flexible and easier-to-use platform to build, test and deploy applications, one that is now also becoming more secure. This is vital for the development of 5G services at a time when the use cases for 5G are still being defined. Operators will need to be in a position to spin up services as and when needed to support different use cases; by using containers, it will be possible to serve customers quickly and efficiently.

Another key aspect is the need to deliver services and applications closer to the end user by utilizing mobile edge computing. This is integral to ensuring the low latency and high-bandwidth associated with 5G and will support use cases across a wide range of verticals including transportation, manufacturing and healthcare. However, flexible architectures will be required to support this type of infrastructure throughout hybrid cloud and virtualized environments. As operators move network infrastructure to the edge, the use of containers will become pivotal to supporting 5G applications. 

The use of microservices and containers will increase during 2019 as operators ramp up their 5G propositions. Despite offering clear advantages, they will also add a new layer of complexity, and carriers will need to have clear visibility across their service delivery network if they are going to make a success of 5G.

Operators will ‘scale or fail’ to meet the 5G demand in 2019

5G will be faster, smarter and more efficient than 4G, but in order to meet demand and to support new architectures, networks will have to scale. While most of the scale in the core network will be cloud and software-based, there will still be a need for hardware and equipment at the network edge—and in a 5G environment there will be a lot more equipment. In fact, the number of cell sites will increase dramatically to support and propagate the higher-frequency bands that will transmit 5G data traffic over the air. This is when network management tools will come into their own. In 2019 we will see the deployment of automated networks driven by software and controlled by virtual machines and artificial intelligence.

Network automation and orchestration are by-products of virtualization and will add another layer of complexity. However, they are also integral to the rollout and sustainability of 5G networks, particularly as network topologies will change to accommodate a combination of small cell and macro cell sites. Small cells in particular will form the bulk of the new RAN (radio area network), and they are expected to increase cellular networks threefold. If network engineers think they already have enough issues to deal with today in maintaining 4G/LTE networks, then they may be in for a shock as 5G networks are gradually rolled out. In fact, without having total visibility of these more complex and expansive networks, 5G in the RAN is going to become extremely difficult to manage.

If the number of cells were to double or triple, not only would network engineering teams need to have the full confidence in their network management tools to make sure the network is running optimally, but they would also be faced with one heck of a job troubleshooting hundreds, and potentially even thousands, of cells if an issue arose.

In 2019, carriers will be scrutinizing costs per cell site as they look to invest in new infrastructure. They will look to offset any costs by implementing intelligent and automated systems that can support 5G networks. However, carriers need assurances that these systems are providing them with the right information about the uptime and performance of their new networks. Having a window into this multi-layered and virtualized environment—and being able to extract smart data in near real-time—will be essential for the ongoing management of new 5G networks. Indeed, the only way to achieve the requisite level of automation will be to have complete visibility of these complex new architectures with real-time smart data that can inform policy engines in a closed feedback loop.