Data management, distribution, and processing for the next generation of networks

Introduction

Mobile and fixed networks are transforming. Anyone who follows progress in the rollout of 5G networks will be familiar with factors such as cloudification, containerization and automation, which are transforming network design, build, and operations. At the same time, fixed networks will also benefit from many of the innovations that will transform the mobile domain. F5G – Fixed 5G – creates a core, common approach to both.

In this context, many Communications Service Providers are also forging new alliances and partnerships with stakeholders from the big tech and hyper-scalers. While these will help CSPs on their journey towards new levels of elastic agility, they also bring challenges and opportunities for combining new data sources with traditional network information.

Similarly, every stakeholder in the extended ecosystem – CSPs and hyper-scalers alike – is also confronted with the challenge (and promise) of an emerging form of Artificial Intelligence: Generative AI.

Top of our list though, is data democratization. As CSPs spread data access and hence insights across their organizations, moving from experts to casual users. So, what does this flux herald for 2024? Read on to learn what Elisa Polystar’s experts forecast for the next 12 months.

Open data architectures drive data democratization

Access to data is key to spreading insights across an organisation and to equip stakeholders to make better decisions. Typically, access has been restricted to expert user groups and has been siloed in different tools and platforms.

However, with a truly open data architecture, data from any relevant source can be aggregated into a single platform, where it can be consolidated and mediated – and made accessible to any interface through which it can be interrogated. This can include both tools already in use in the organisation, as well as new platform and search functions. However, given the volume of data – and the variety of available sources – a prerequisite for such a platform is scale, which will likely require the involvement of hyper-scale cloud solutions.

But, by decoupling data from complex silos, the consumption of such data can be spread as a discipline across the organisation, at all levels, breaking the monopoly of data scientists and domain experts. Expect significant moves towards such data democratisation programmes in 2024.

Generative AI takes lead over Predictive AI

Predictive AI has entered the telecoms domain. It has delivered benefits across all network layers, supporting proactive service assurance and automation efforts. However, Generative AI has been unleashed, with a new level of maturity – in 2024, we can expect it to deliver an abundant harvest of new fruits.

Why? It’s really about accessibility and interrogation. ChatGPT, Scribe, AlphaCode, GitHub Copilot and the like provide easy-to-use interfaces that enable interaction with AI algorithms trained on large data sets. Put simply, users can ask semantic questions and obtain answers, generated from the available data. So, humans can discover insights and obtain answers that are structured in a way that they can understand.

In the CSP domain, this has huge potential to enhance customer service, for example. Instead of reading how-to guides, complex user manuals, and so on, when presented with a Generative AI tool, anyone can ask questions like “how can I adopt an eSIM”?

By using the right data sets, CSPs can offer enhanced and rapid support, relieving a significant burden from customer service teams. They can also use Generative AI interfaces to triage questions. Instead of waiting in a queue, calls can find answers to common questions – through CSP support portals or apps. At Elisa Polystar, we also expect Generative AI to make an equally significant breakthrough internally.

By using CSP data, Generative AI can open a window on insights specifically from a CSP’s network. The same information that is, today, processed and presented through dashboards and SQL query interfaces to experts, can also now be accessed via simple interfaces.

Internal stakeholders can ask questions like “how many call drops were there in Stockholm last week?” or “which cell sites have had the least amount of traffic in the last six months?” or even “how many of my 5G subscribers are using 5G radio more than 50% of the time?”

Such insights will enable CSPs to deliver a wider set of KPIs and data, to more users – helping drive efficiency and to enhance optimisation efforts.

Federated cloud and edge computing finally becomes a reality

Investments in cloud architectures by CSPs span their own resources – but they also extend to third parties; federated cloud architectures are the result. These interconnected cloud assets allow CSPs to extend their reach, share resources and collaborate with other stakeholders to secure desired outcomes.

Why do we combine this with edge computing? Because resources at the edge may not be in the CSP’s own domain. Edge systems may be a combination of CSP-owned and other resources that are used in parallel to deliver a particular service. And, regardless of overall pace towards 5G SA, edge computing is now firmly in demand by enterprises (and CSPs), to support a new generation of high-performance and low latency services.

This demand won’t only be served by CSPs, however. Many enterprises are seeking to deploy private networks – and the resources required to support their applications may be accessed via federated clouds. This user may not need its own UPF, but it may benefit from one offered by another provider in an adjacent edge location, or delivered by a systems integrator that runs multiple private networks with shared resources, available on demand.

Investment in evolution plateaus?

The pace of network evolution does indeed seem to have slowed. 5G SA deployments haven’t followed the trajectory anticipated a couple of years ago. Of course, rolling out a full 5G SA network on a national basis is undoubtedly costly. It’s not just the tech stack, there’s a lot of other kit required. However, is this situation set to continue?

If the primary vehicle for 5G monetisation is enterprise services, then a good indicator that things are picking up is the surging interest in 5G non-public networks (which are almost exclusively for enterprise purposes, whether delivered by CSPs or others.

What matters here is that some in the industry await a 5G killer application. This is misguided; there isn’t a single application that will dominate, rather an expanding range of specialised applications enabled by full 5G architectures that will be demanded by enterprises of different kinds and in different sectors.

So, growth may have plateaued, but we expect this to be temporary as the industry realigns behind enterprise opportunities – whether through dedicated networks, or as slices or hybrid networks from national network investments.

Conclusion

While widespread adoption of 5G SA has indeed slowed, it remains a clear goal for CSPs, in particular because of the enterprise revenue it is expected to unlock. Edge computing is gathering pace – a promising signal. But, in parallel with these, it is clear that data and how it can be accessed is central to future operational transformation and optimization efforts. Not only is data democratization likely, it is also a prerequisite for ensuring that operational agility and efficiency goals can be reached – and for delivering the promise of new service capabilities.

Download article to read later

This article was originally published on Fastmode – Trends & Predictions.