Predictions 2021 – Automating for the cloud, the end of the NOC and more

Mar 8, 2021 | Editorials

2020 was a year of unprecedented change. Of course, much of this change was forced upon the industry, as we adjusted to an entirely new set of circumstances. The impact of this will be felt for years to come and will accelerate several shifts that were already underway – but it will also ease the transition to a new level of network performance that demands profound changes in operational systems. So, what’s in store for 2021?

Relentless shift to the cloud and a hyper-distributed edge fabric

Let’s start with the cloud. Operators have been exploring cloud processing and virtualisation for a number of years, partly to reduce costs and partly to drive new levels of performance. Some have invested in their own cloud infrastructure, while others have sought to partner with cloud providers. Many are exploring hybrid models. This change is now irreversible and 2021 will see an acceleration of the different cloud models. Here’s why.

5G core networks are cloud native by design and deployed in data centres, privately owned or otherwise. They are also designed to deliver a wider range of services (beyond mobile broadband), supported by network slicing. These services and slices may be subject to stringent Service Level Agreements (SLAs). As such, the way networks are managed will change completely.

In many new cases, intelligence will shift to the edge. That’s because the huge opportunity for 5G is generally understood to come from enterprise and vertical industry needs. Such stakeholders want to capitalise on low latency performance, increased network availability, and new wireless speeds, to deliver new applications and services.

As a result, network connectivity and content must move towards devices. To support this and to capitalise on new enterprise business opportunities, operators will extend their cloud from the core to the edge and also to enterprise domains. They must manage services that utilise network functions in private, public or enterprise clouds, backed by the interaction of new hyper-distributed cloud assets. To achieve this, operators will be dependent on a mix of assets. 2021 will see a slew of cloud investments, as operators shift to embrace these opportunities.

AI for networks takes off

In this environment, operators will have to analyse vastly more data, from multiple domains, to understand the experience of their customers. There’s just too much of this to process without help, so 2021 will be the year that new data processing techniques, backed by Artificial Intelligence and Machine Learning really take off.

AI and ML will enable operators to process new sources of data at vast scale. We’ve already seen extremely promising results from different use cases. For automation with AI and ML allows anomalies to be detected faster, without human intervention and with fewer errors. ML enables software to learn, so that results can be refined and enhanced, leading to continual process enhancement.

Operators are gradually adopting these innovations, and this will accelerate in 2021, because what were once novel technologies are now showing that they can deliver benefits, at scale, providing measurable outcomes that enhance operations while meeting strategic goals.

The end of the NOC

The Network Operations Centre (NOC) has been fundamental to the way in which operator networks are managed for decades. However, it requires considerable resources and concentrates these into centralised facilities. The NOC is ripe for automation and solutions are now available that enable this to be achieved. While some have adopted an outsourcing model, placing their operations in the hands of partners, this approach has also been exposed as a source of risk during the pandemic.

Insourcing is now back – but rather than simply staffing the NOC with their own employees, operators will adopt automation solutions that enable them to take control, remove external dependencies, and allow them to focus resources on performance enhancements and new service revenue. Several leading operators have already taken this step. This trend will soar in 2021.

Adapting to new Service Level Agreements

SLAs are key to contracts between operators and enterprise customers. They govern how services are delivered, define performance targets and set out how remedies can be sought when they are not met. 5G brings a host of new services, particularly in the enterprise domain. These will have a much wider range of performance requirements.

Operators will have to create, support and deliver a huge number of new SLAs, so must ensure that they have the operational systems in place to monitor their performance and take appropriate steps to enforce SLA targets. This will require new forms of real-time monitoring, analytics and reporting, with automation to ensure performance optimisation.

Key to this will be close alignment between orchestration and service platforms, and the solutions that monitor them and enable optimisation and remedial action to be taken. Closed loop actions will be essential.

Expect 2021 to see operators gear up to embrace this challenge and to deploy new automation solutions that will support their efforts to monetise new 5G edge and enterprise services – and deliver these new SLAs.

Introducing the Network Data Analytics Function

The Network Data Analytics Function – or NWDAF –was introduced in 3GPP’s R15. R16 and R17 have built on this and we expect the NWDAF to gain attention this year.

That’s because the NWDAF has become central to solving the problem of managing highly distributed, multi-vendor networks. As we noted, there’s a clear need to perform real-time analytics to support delivery of new SLAs, process automation, and network performance optimisation across a plethora of new services and for a growing number of new opportunities.

To do that, data must be collected from different systems and made available to requests from supervisory systems and operational processes that make use of the 5G core network functions.The collection of such data has historically relied upon solutions that can interface with proprietary platforms, often operating in different silos. The NWDAF removes this dependency by providing a central resource that interfaces to all network functions – including those at the edge – and provides analytics as a service when requested by other processes that are part of the service framework.

It leverages standards-based interfaces from the service-based architecture (5G SBA) to collect data and provide it as required, at the right scale and in real-time. It’s a key component in the move towards new levels of operational performance and in support of 5G services.


You might have noticed a common theme here. The cloud, AI / ML, the NOC and the NWDAF are tools that will help operators deliver on the promise of 5G. There’s huge excitement in the industry concerning the potential of 5G, but we haven’t (yet) seen all of the promised benefits. In some quarters, optimism has turned to skepticism. Can 5G really do what was claimed?

The answer is yes – but further steps have to be taken to provide the supporting infrastructure and new processes before that can happen. Rest assured, with the realisation of these predictions, 5G will deliver, at last, and this will be seen in the new, more complex and diverse SLAs that result.

This article was originally published in FastMode