The DTW24 - Ignite event app is now live!

Registered attendees can now download the app on your iOS, Google Play device, or access the app via browser.


Sense, Think, Act: Three Essentials for Autonomous Networks

By: Kal De, SVP Product and Engineering, Cloud and Network Services, Nokia

Today’s consumers and enterprises expect and demand services that can be consumed in a frictionless manner. That means that telcos need to deliver services with Zero Wait (no delay between order and fulfillment of the service), Zero Touch (delivered automatically without the need for human intervention), and Zero Trouble (flawless quality throughout the service experience).

TM Forum’s Autonomous Network Operations provides a framework for telcos to achieve ‘Network Nirvana’ – that is, Level 5, fully autonomous operations. Nokia believes that to reach this goal we need to enable networks that can sense, think, and act.

Sense: new approaches to tame cloud complexity

Telecoms have always been complicated, and the move towards cloud-native architectures and the consequent disaggregation of networks has exacerbated this complexity. Network functions are now split into multiple microservices, running in distributed containers across different cloud environments (private, public, hybrid). How can telcos effectively troubleshoot network issues across cloud infrastructures, platforms, and applications? How can they manage the performance of this complex environment to deliver Zero Wait, Zero Touch, and Zero Trouble services?

New approaches are needed to allow telcos to accurately ‘sense’ the dynamic cloud environment that underpins service operations. Embedding observability into service assurance systems is a good starting point to deliver contextual understanding and ensure that the OSS view of the world matches reality. Unified inventory systems are another key building block, providing a real-time view of all network resources, topologies, and services.

Think: improved data quality for better decisions

AI is essential to Autonomous Operations, enabling networks to ‘think’ and make intelligent decisions. But without a continual stream of high-quality data to fuel them, even the most sophisticated AI algorithms are useless. The extreme diversity of telecom data compounds the challenge, leaving CSPs to grapple with an unprecedented variety of data sources and vendor-proprietary formats that make effective governance a nightmare. Indeed, an Analysys Mason survey showed that 56% of Telcos viewed data quality as a significant challenge.

How could Telcos improve data quality? A data mesh architecture can replace hard-to-manage data to lakes and help generate carefully curated data products. Exposure of these data products via APIs to a catalog would allow data scientists to create new AI use cases much faster.

Better quality data would also increase the accuracy of both traditional machine learning and Large Language Models (more on the role of GenAI later). In turn, this will provide Telcos with the intelligence to rapidly identify anomalies – analyze the root causes of network issues – and prevent future service-impacting degradations from occurring.

Act: closing the loop to automate at scale

Telcos need to act efficiently on the insights provided by AI to deliver Zero Wait, Zero Touch, and Zero Trouble services. Closed-loop automation is required to achieve this at scale: speeding up responses to customer orders, removing friction, and ensuring flawless quality in service delivery. But this is far from trivial: siloed tools, legacy processes, and ‘change fatigue’ are among the factors holding Telcos back from achieving higher levels of automation.

Network Slicing is a good illustration of the value of Closed-Loop Automation. Consider the example of an airport that requires a dedicated slice to support baggage handling, staff communications, and other services. We know that airports can be chaotically crowded, or eerily empty. Demand on the network slice will naturally vary enormously over time, in line with the number of flights and passengers transiting through.

A manual approach to design, deployment, and assurance would be very inefficient, and unable to dynamically scale up or down to optimize network resource usage. Telcos such as Telstra, stc, and Telenor have already applied automation to network slicing to cope with similar challenges.

What’s next for Autonomous Network Operations?

You probably won’t be surprised to know that I believe Generative AI will play a key role in advancing autonomous operations and enabling networks that sense, think, and act. At DTW24 - Ignite, Nokia will showcase examples of GenAI in action.

These will include enhancing the detection and resolution of cybersecurity threats; augmenting the work of network operations teams; removing friction from ordering by automatically translating business intent into service templates – and much more.

I’m super excited to learn for myself how our industry plans to exploit the enormous potential of GenAI and look forward to meeting many of you there. See you in Copenhagen!

About the Author

Kal De leads Products & Engineering for Nokia’s Cloud & Network Services business. His teams build cloud-native software to help CSPs and enterprises create networks that sense, think and act. Prior to joining Nokia Kal held a variety of commercial and product development leadership roles at companies such as Vista Equity Partners, VMWare, Docker and Oracle.

Sponsored by:

Nokia 600x100.png