The mobile communications industry has been provided with new guidance on the building blocks needed for autonomous networks that encompass artificial intelligence (AI) thanks to the publication of NGMN’s “Automation and Autonomous system Architecture Framework – Phase 2.”

The latest updates provide guidance and direction on the use cases, requirements and architectural principles to embrace interoperable, multi-vendor and standards-based autonomous networks that will help to manage the increasing complexity of the evolving 5G Advanced ecosystem.

“AI has huge potential to revolutionise the way we plan, build, operate and maintain networks, the services we offer and our ability to respond rapidly to customer needs,” said Arash Ashouriha, chairman of the NGMN Alliance Board and SVP Group Technology at Deutsche Telekom. “This publication provides a further foundation for the mobile industry to move forward together by developing and embracing a standards-based approach to autonomous networks.”

This framework outlines architectural considerations for enhancing operational capabilities through AI and machine learning (ML), enabling operational capabilities beyond typical human response times amid growing networking complexities, while guiding standards development organisations in delivering a more customised, user-centric experience.

“With the ever-increasing complexity of networks and the move towards disaggregated and cloud-based architectures, it is more fundamental than ever that we manage that complexity by introducing AI at all levels of our networks and operations,” said Michael Irizarry, member of the NGMN Alliance Board and executive vice president and chief technology officer, UScellular. “UScellular is proud to have led this activity working alongside our partners at NGMN. We strongly encourage the industry to adopt this guidance when developing the industry standards that are necessary to ensure autonomous networks can be delivered through open and interoperable standards.”

The autonomous system architecture framework is intended to serve as guidance in the development of interoperable and market enabling specifications, for the continuing advancement of an automated and self-adaptive 5G ecosystem of heterogeneous access, virtualisation, forward-looking service enablers and emerging usage scenarios.

Topics covered in the publication include:

• Management of complexity
• Network, service and device evolution
• Autonomous System Management and Orchestration and the role of Large Language Models (LLMs)
• Co-operative technologies and processes
• Use cases
• Security and privacy
• Role of industry standards.