Deployment of the 5G core is the next big step in 5G networks. It will enable standalone (SA) operation and new service types. This new core will be deployed on software-defined infrastructure and should be automated by design.

Automation of the 5G core is required. First, because operators do not yet know which services will take off and achieve commercial success at scale and which will emerge as valuable but niche. And second, because operators must reduce the cost of production to remain competitive and support a diverse services portfolio.

The new "Heavy Reading Cloud-Native 5G Core Operator Survey" published in March 2021 shows operators will be purposeful and enthusiastic in their adoption of 5G core automation. However, they will not be "gung-ho" in its implementation.

This assessment is illustrated in the figure below. The question asks how quickly operators expect to be able to make changes to the topology or services running across the 5G core. The response reveals that one-week, or less, cycle times are expected to be sufficient for core network topology changes, with 70% of the respondents selecting this "balanced" view, ahead of just 15% selecting the more "aggressive" daily update option.

Figure 1: How quickly would your company expect to be able to roll out changes to the topology or services going across your 5G core? n=72
(Source: Heavy Reading)

A one-week cycle time is much faster than is typical in today's 4G mobile core network and will not be achieved in 5G without tools that automate infrastructure, application deployment and service lifecycle management. The good news is that advanced operators are now close to a position where they can technically make frequent changes. The "one week or less" target is now within sight, even if it is not yet commonplace.

For good reason, operators are also cautious about 5G core automation. In a core network, even small errors can result in a widespread network failure, which is a problem not only for customers and the operator, but also (potentially) for society at large. As more devices and services – and more critical services – run on 5G networks, the impact of outages and brownouts becomes ever more severe. Operators will inevitably, therefore, be cautious in how fast, and how far, they rely on automated operation of the 5G core network.

So, does this one-week cycle time for changes to the 5G core represent more of an aspiration than reality for operators?

Speaking on the Light Reading "5G Core Operator Survey Webinar" (view the archive here), Richard Band, head of Mobile Core and 5G at Hewlett Packard Enterprise (HPE), says not. "The desire is really there," he says, "but automation needs to be built-in from day one."

"Significant effort is required to achieve this balanced, but still aggressive, objective of making changes in less than one week," says Band. "Working with customers, we see exactly this: two days to validate, and one day to rollout, is the target. So, the one-week or less [cycle time] is really a true objective."

The Heavy Reading Cloud-Native 5G Core Operator Survey focuses on the 3GPP 5G core, associated cloud infrastructure platforms and end user services. You can download a PDF copy here.

— Gabriel Brown, Senior Principal Analyst, Mobile Networks, Heavy Reading

This blog is sponsored by HPE.