Context-Based e2e Autonomous Operation in B5G Networks.

Area of Science:

• Telecommunications Engineering
• Network Architecture
• Artificial Intelligence in Networking

Background:

• Fifth-generation (5G) networks established foundations for smart radio access networks (RANs) and autonomous fixed network operation.
• Beyond 5G (B5G) and 6G networks require smart end-to-end (e2e) operation for stringent service demands.
• Coordination between smart RAN slice management and autonomous fixed transport networks is crucial for cost and energy savings.

Purpose of the Study:

• To propose a novel context-aware network operation approach for B5G/6G.
• To enable seamless coordination between RAN slices and the fixed transport network.
• To ensure reliable and efficient autonomous operation of optical connections supporting RAN slices.

Main Methods:

• Developed a context-aware network operation concept.
• Utilized aggregated and anonymized slice operation information sent asynchronously to the fixed network orchestrator.
• Employed artificial intelligence (AI)-based models for predictive capacity management of optical connections using slice context.

Main Results:

• Slice context availability significantly improved predictive methods, reducing prediction maximum error by 90%.
• Achieved robust and efficient optical network operation for dense RAN cells (>32 base stations per cell).
• Demonstrated superior performance compared to benchmarking methods in heterogeneous service demand scenarios.

Conclusions:

• Context-aware network operation is vital for accurate and reliable autonomous fixed network operation under dynamic traffic.
• The proposed approach ensures that the benefits of smart slice operation are not negated by uncoordinated fixed network operations.
• This method enhances the efficiency and robustness of optical networks supporting future mobile network generations.