An Agile Conflict-Solving Framework for Intent-Based Management of Service Level Agreement

Managing and orchestrating Radio Access Networks (RAN) are error-prone tasks due to the high number of parameters and the complex interaction among all the network components. This paper proposes the AGIR system (AGility in Intent-based management of service-level agreement Refinements) for Open Rad...

Full description

Saved in:
Bibliographic Details
Published in2023 2nd International Conference on 6G Networking (6GNet) pp. 1 - 8
Main Authors de Oliveira, Nicollas R., Moraes, Igor M., de Medeiros, Dianne Scherly V., Lopez, Martin Andreoni, Mattos, Diogo M. F.
Format Conference Proceeding
LanguageEnglish
Published IEEE 18.10.2023
Subjects
6G mobile communication
Cellular Communications
Measurement
Mobile Networks
Next-Generation Networks
Proposals
Real-time systems
Scalability
Training
User experience
Online AccessGet full text

Cover

More Information
Summary:Managing and orchestrating Radio Access Networks (RAN) are error-prone tasks due to the high number of parameters and the complex interaction among all the network components. This paper proposes the AGIR system (AGility in Intent-based management of service-level agreement Refinements) for Open Radio Access Networks (Open RAN). AGIR implements intention-based network management, allowing operators to specify Service-Level Agreements (SLAs) for the RAN to fulfill. The system translates imprecise intentions into configurable network instructions, enhancing flexibility, scalability, and reducing human errors. The proposal enhances productivity, lowers operating costs, improves user experience, and optimizes network performance through real-time data analysis and automation. AGIR addresses the need for open and comprehensive interfaces in RANs, enabling cooperation among components from different suppliers. Our results reveal that the proposed system reaches more than 80 % precision in detecting conflicting intentions when deploying a deep neural network based on Long Short-Term Memory with 256 neurons.
DOI:10.1109/6GNet58894.2023.10317714