How programmable networks with automation and AI can enable differentiated connectivity and monetization of 5G services

By: Miriam Deasy, Senior Solutions Marketing Manager, Ericsson; Jaime Trapero, Business Area Cloud, Software & Services Marketing Director, Ericsson; David Espadas, Global Marketing Director for Network Automation and AI, Ericsson

The telecommunications industry is facing unprecedented challenges and opportunities in the 5G era. The demand for connectivity is growing exponentially, driven by new use cases, devices, and applications. Network complexity is also increasing, as service providers need to manage multiple layers, domains, and technologies, in an increasingly multivendor environment. Differentiated connectivity can drive a differentiated user experience by programming network capabilities end-to-end to address specific use cases, instead of just the single best-effort.

Creating value out of 5G standalone (SA) capabilities, and other associated technologies, enables connectivity service offerings that can be used to tailor connectivity. With this, CSPs can offer connectivity in a way that achieves network system efficiency and enables scalability and differentiated pricing models.

To cope with these challenges and to capture the opportunities, service providers need to transform both networks and operations with automation and artificial intelligence (AI).

Autonomous, programable networks are the goal of this transformation, where the networks can self-configure, self-optimize, self-heal, and self-learn, without human intervention. The industry is coalescing around a common view of the journey towards autonomous networks, with five levels of autonomy defined. Level 4 is the next aspirational state, where the network can handle complex tasks and scenarios, and where humans only supervise and set policies.

But where are we on the capabilities required to reach level 4? And what are the benefits of autonomous networks for service providers and their customers?

The drivers and challenges for autonomous programmable networks

Autonomous networks can reduce the OPEX and improve the network performance by automating repetitive and complex tasks. Enable service differentiation and monetization by providing customized and dynamic connectivity for different use cases. However, achieving autonomous networks is not a trivial task. It requires a significant upgrade in the foundational capabilities of the network and the operations, such as end-to-end orchestration, AI and automation, open APIs for exposure and intent-driven operations.

The foundational capabilities for autonomous programmable networks

End-to-end service orchestration and assurance enables network slicing, which is the most advanced form of differentiated connectivity, where each slice can have its own performance, quality, and security characteristics. Dynamic network slicing can also enable the network to adjust the slice parameters and resources in real time, based on the changing demand and conditions.

One of our customers, Telstra, is a pioneer in network slicing service orchestration, leveraging Ericsson Dynamic Network Slicing solution, rooted in Ericsson Service Orchestration and Assurance. Telstra successfully orchestrates wireless services for network slicing across multiple domains, such as radio, transport, and core, and achieved advanced network slicing capabilities for Australian enterprises.

AI and automation enhance the network performance, efficiency, and quality, in every network domain. They can help to automate and optimize complex tasks, such as load balancing, interference management, traffic steering, and anomaly detection. They can also help to provide better customer experience, by predicting and preventing issues, and by personalizing the services and offers.

Explainable AI can help to ensure the trustworthiness and explainability of the AI models and decisions, especially when they have a direct impact on the network and the customers.

In the RAN domain, the new paradigm for AI and automation is embodied in the Service Management and Orchestration (SMO) and the RAN Intelligent Controller (RIC), which enable a more granular and dynamic control of the radio resources and services.

The SMO architecture is already empowering a rich ecosystem of rApp developers that will in the coming years increase innovation and freedom of choice for CSPs in this area. Ericsson’s Intelligent Automation Platform ecosystem is an example of this.

To maximize the benefits from the new services underpinned by programmable networks, a new approach is needed for service exposure via APIs that offers flexible support for business models ranging from direct relationships between the CSP and the consumer to indirect ones facilitated by aggregators like Vonage.

Intent-driven operations: the key to building autonomous networks

Intent-based operations (IBO) allow the service providers to express the desired outcomes or goals of the network and the services, and to let the network optimally manage itself to achieve them. A cognitive loop that leverages AI, machine learning, and automation can simplify service definition and network operation, particularly within the dynamic 5G environment.

One of our customers, Digital Nasional Berhad (DNB), is a frontrunner in leveling up their network operations from data-driven to intent-driven, leveraging on our Ericsson Operations Engine solution. DNB created the world’s first multi-operator core network, which met the needs of six different telecom service providers and ensured affordable 5G connectivity for the people of Malaysia, just six months after its founding.

DNB also pioneered on supporting multiple service providers by guaranteeing premium service-level agreements (SLAs) and therefore unlocking the monetization of new services by consistently delivering differentiated connectivity services through network slicing.

Figure 1 What is the Autonomous Networks Journey.png

Intent-based operations are also the test bed and reality check for autonomous networks. They allow the service providers to experiment with different levels of autonomy and to evaluate the value of the automation and AI solutions. Rather than aiming for a specific level of autonomy, service providers should focus on the value that each level of autonomy can bring to their network and customers and adopt the solutions that best fit their needs and goals. This is the essence of the "don't chase the badge, chase the value" idea.

Conclusion

The journey towards autonomous programmable networks is underway, and many of the required technologies are available to advance towards level 4 with a 2030 horizon. But these technologies are not plug-and-play. They require a transformation mindset and a holistic approach across networks and IT, to ensure they are deployed and integrated cohesively and effectively. Autonomous networks will be as revolutionary as the new possibilities for differentiated connectivity and monetization that they will enable. Are you ready for the ignition and liftoff?

If you want to hear more from our customers and partners on this topic, join us at the DTW24 - Ignite, where Ericsson is delighted to be a Platinum Sponsor. We will showcase our latest solutions and innovations for autonomous networks.

You can also check out our blog posts, case studies, and e-briefs on our website, where you can learn more about our vision and offerings for end-to-end service orchestration and assurance, AI and automation, and intent-driven operations. We look forward to seeing you at the event and hearing your feedback and questions.

More from our customers on this topic at DTW24:

Telstra - Talking about the evolution of service orchestration and assurance and exposure, on Tuesday 18 June, 16:00, Stage B
Swisscom – Talking about how AI is revolutionizing customer experience and what is coming for AI in Telco, on 18 June, 15:00, stage D
DNB – Talking about how intent driven autonomous networks deliver premium service levels and desired business outcomes, on 18 June, 11:30, stage C

Learn more: