please do not alter the heading in the table below as it is used for reporting purposes |
50 Attendees
See TMF-IETF standardization progress-Xing.pdf
See ITU-T_SG13_AN activities_update-Marco Carugi.pdf
See ETSI_OCG_ANMSDOLuigi29425.pptx
See ETSI ISG F5G AN updates v2 (Yi Lin).pptx
Call for IPR Declaration
Informational Meetings (Open Meetings) “This is an open information meeting in which no information or materials furnished or provided by the attendee shall constitute or contain intellectual property and will not be treated or protected as such. All attendees understand and accept this statement.” |
Meeting Type | Informational (Open) Meeting |
---|---|
IPR Claims | N/A |
Topics | Presentations |
---|---|
| |
1. Zhaoxing (CAICT) Topic: IETF AN Standardization: Network Management Agents Driving L4+ High-Level Autonomous Networks | TMF-IETF standardization progress-Xing.pdf |
2. Marco Carugi (ITU-T SG13 Q20 Rapporteur) Topic: Update on AN Activities in Progress within ITU-T SG13 (Q20) | TU-T_SG13_AN activities_update-Marco Carugi.pdf |
3. Luigi Licciardi (Huawei, OCG AN) Topic: AI for the Evolution of Autonomous Networks | ETSI_OCG_ANMSDOLuigi29425.pptx |
4. Yi Lin (Huawei, F5G Technical Manager) Topic: Autonomous Network Levels Classification and Evaluation for Fixed Networks | ETSI ISG F5G AN updates v2 (Yi Lin).pptx |
IETF AN Standardization Progress: Xing Zhao from CAICT discussed the status of IETF autonomous network standardization. He highlighted the progress made in defining concepts and frameworks for autonomic networking and intent-based networking. Xing also mentioned the introduction of AI into network management and the challenges that need to be addressed. He presented a draft on AI-based network management, which proposes a standardized common architecture for AI network management. The draft aims to provide a framework for AI network management, clarify its relationship with existing controllers, and define functional requirements for different scenarios. Xing concluded by stating that the standardization of AI agents in IETF is still in its early stages and needs to be accelerated in the future. Xing presented on the ongoing discussions about network management and the importance of natural language in the process. Jörg raised questions about the realization of the intent northbound API and the role of natural language in network management. Wadii suggested the use of reasoning capabilities in LLMs for troubleshooting and network management.
Update on AN Activities in Progress within ITU-T SG13 (Q20): Marco provided an update on the activities of Study Group 13, focusing on AI and ML. He highlighted the group's work on autonomous networks and the integration of AI capabilities in networks to make them more efficient. Marco also mentioned ongoing projects, such as the development of a knowledge management system for autonomous networks and a focus group on native telecom networks. He emphasized the importance of collaboration and the potential for reciprocal collaboration with other SDOs. Marco also mentioned a correspondence group on data sets, which aims to provide guidelines for using data sets for ML in networks. Marco and Paul discussed the architecture framework for autonomous networks, addressing questions about its compatibility with other architectures. They explained that the framework is conceptual and aims to create control loops for network adaptation. They also touched on the topic of trust in AI models, emphasizing the need for a framework to assess model trustworthiness. Jörg raised concerns about the lack of a framework for verifying AI model quality and correctness. Paul mentioned ongoing work on explainability and trust services in AI networks. The team agreed on the importance of collaboration on these topics.
AI for the Evolution of Autonomous Networks: Luigi discussed the coordination group, OCG, which facilitates the exchange of results and deliverables for autonomous networks. He highlighted the importance of identifying synergies, best practices, and common requirements. Luigi also mentioned the role of artificial intelligence in autonomous networks and the need for continuous dialogue with the European community. He presented a white paper on the evolution of autonomous networks, focusing on the role of AI and the challenges and opportunities of level 4 evolution. The white paper is scheduled for release on July 25.
Autonomous Network Levels Classification and Evaluation for Fixed Networks: Yi discussed the Fixed Network Autonomous Level Classification and Evaluation based on the TMF. He explained the concept of F. 5G, the 5th generation of fixed networks, and how it has evolved over time. Yi also introduced the Intelligent FTTR (Fiber to the Room) and AI for Optical Networks white papers. He emphasized the importance of cooperation between different standard organizations for the development of autonomous networks, particularly for the fixed optical network. The conversation ended with the proposal of two more meetings scheduled for July 29th and October 28th.