NMRG 48th meeting IETF 102, Montreal, Canada RG Chairs: Laurent Ciavaglia Lisandro Granville Available During Session: Slides: https://datatracker.ietf.org/meeting/102/session/nmrg Meetecho: http://www.meetecho.com/ietf102/nmrg/ Available Post Session: Audio recording: https://www.ietf.org/audio/ietf102/ Video recording: https://www.youtube.com/user/ietf/playlists *********** * Session 1 Thursday, July 19th, Afternoon session I, 13:30 – 15:30 (2 hours) Room: Duluth Agenda: 1. Introduction, Chairs 13:30, 10 min. 2. Topic: Telemetry & Co [13:40 – 14:10] 2a. Exploiting External Event Detectors to Anticipate Resource Requirements for the Elastic Adaptation of SDN/NFV Systems, Pedro Martinez-Julia https://datatracker.ietf.org/doc/draft-pedro-nmrg-anticipated-adaptation/ 13:40, 10 min. 2b. Toward a Network Telemetry Framework, Haoyu Song https://datatracker.ietf.org/doc/draft-song-ntf/ 13:50, 10 min. 2c. Q&A, discussion and way forward 14:00, 10 min. 3. Topic: Management aspects of network slicing [14:10 – 14:40] 3a. Management slicing problems and gap analysis, Alex Galis 14:10, 10 min. 3b. (remote) Network Slice use cases, Kiran Makhijani 14:20, 10 min. 3c. Q&A, discussion and way forward 14:30, 10 min. 4. Topic: Novel approaches to network management [14:40 – 15:30] 4a. Intelligent Network Management using Reinforcement Learning, Yong-Geun Hong https://datatracker.ietf.org/doc/draft-kim-nmrg-rl/ 14:40, 10 min. 4b. Roadmap to a Networkless World, Bing Liu https://datatracker.ietf.org/doc/draft-liu-nmrg-networkless-roadmap/ 14:50, 10 min. 4c. (remote) Autonomics 3.0, Jéferson Campos Nobre 15:00, 10 min. 4d. Distinguishing Intent, Policy, and Service Models, Alex Clemm https://datatracker.ietf.org/doc/draft-clemm-nmrg-dist-intent/ 15:10, 10 min. 4e. Q&A, discussion and way forward 15:20, 10 min. 5. Stay tuned for the Special Meeting on AI/ML in Networking! Same room, at 18:10 *********** * Session 2 Thursday, July 19th, Afternoon session III, 18:10 – 19:10 (1 hour) Room: Duluth IRTF Special Meeting on AI/ML in Networking In the last few years and especially in the last few months, the use of Artificial Intelligence and Machine Learning in networking went from a nice future application to an almost must have. Diverse networking operations from trend detection in traffic monitoring, beam assignment in 5G, congestion control and intent-based networking have identified AI as a means to improve their overall performance under more and more dynamic conditions. One may also imagine how AI can impact the very design of networks and protocols, or else how to solve networking problems in original ways using novel and network-specific artificial intelligence techniques. Several special issues and conferences have already been organized under these themes and the community wanting to use these new tools is rapidly expanding. In addition to these, network architects and designers are also considering the impact of the use of AI/ML on the network themselves. For example: AI in vehicles generates very large data sets that need to be sent to the cloud for analysis or to rely on fog/edge computing. Traffic engineering also requires analyzing very large amounts of real-time information that then needs to be sent back to the network nodes for updates and new features. And the application landscape is rapidly expanding each with its own requirements on the network. The meeting wants to bring together the IETF and IRTF community that has interests and stakes in the development of both AI/ML based tools and on the networks that will support them. While we are targeting an open discussion we already have some presentations planned and you are all welcome to submit a request before Friday, July 13th so we can finalize the agenda. Hoping to see you in Montreal! Marie-José & Laurent