ENI#13 Progressing Release 2 and ETSI / ITU-T workshop “ENI-Machine Learning in communication networks”
Due to confinement and travel restrictions, the ENI#13 meeting was successfully organized online only on 17-20 March 2020.
- 32 were present including 8 operators
- 4/5 Government Ministry Institutes are members: China, Japan and South Korea
- 133 documents were handled
A workshop “ENI-Machine Learning in communication networks” was organized on 16 March between two ETSI ISGs, namely ENI and SAI (Securing AI), and ITU-T’s Q20/13 and FG ML5G “Machine Language 5th Generation”, on AI/ML. This workshop was instrumental in enabling synergies between ETSI and the ITU-T in this field and to better understand the needs of industry. ETSI ISG ENI will send the template of ENI Use Cases to the ITU-T groups by liaison for their reference in future work.
As for the meeting, it was very productive and achieved significant progress. After the publication of the approved ENI 006 PoC framework revision, all PoCs will be required to show interworking on an external reference point. The meeting also progressed on the draft reports GR ENI 008 on Intent Aware Network Autonomicity and GR ENI 009 on Data Processing mechanisms. Major progress on the evaluation of categories in Draft GR ENI 010 was made discussing a five dimensional system of quantification of the Classes published last year in GR ENI 007.
Significant progress was made on the learning techniques for AI in the ongoing revision of ENI 005. An open area was also approved for all stable drafts and previously published deliverables. Attendees continued drafting revisions of the next versions on ENI use cases, requirements and Terminology for Release 2.
In addition, we started working on Evaluation of the AI Network Configuration and Mapping of operational systems to ENI architecture. ISG ENI is now progressing into the 2nd Release.
Work in progress of ETSI ISG ENI in Release 2
The Use Cases ENI 001, Requirements ENI 002 and System Architecture ENI 005 are all open for the next Release and work on updates has started accordingly. An update of Terminology ENI 004 was approved as well. New work on big data and its classification was approved.
Experiential Networked Intelligence (ENI) Release 2 specifications and reports will include:
- ENI 001 Use Cases
- ENI 002 Requirements
- ENI 004 Terminology
- ENI 005 System Architecture
- ENI 006 PoC Framework “Approved in the meeting”
- ENI 008 Intent Aware Network Autonomicity
- ENI 009 Data Mechanisms
- ENI 010 Evaluation of categories for AI application to Networks
- ENI 011 Mapping between ENI architecture and operational systems
Status of Proofs of Concepts
Title | PoC Team Members | Main Contact | Start Time | Current Status (March 2020) |
PoC #4: Predictive Fault management of E2E Multi-domain Network Slices | Portugal Telecom/Altice Labs SliceNet Consortium | António Gamelas Rui Calé | Mar-2019 | Ongoing |
PoC #5: AI Enabled Network Traffic Classification | China Mobile | Weiyuan Li | Jun-2019 | Ongoing |
PoC #8: Intent-based user experience optimization | China Telecom / Huawei Technologies China Telecom, Huawei Technologies, AsiaInfo, Beijing University of Posts and Telecommunications | Dong Li | Jan-2020 | Started |
PoC #9: Autonomous Network Slice Management for 5G Vertical Services | Nextworks TIM, Nextworks, Samsung, WINGS, UC3M | Gino Carrozzo | Jan-2020 | Started |
PoC #10: Intelligent Telecom Network Energy Optimization | China Mobile China Mobile Research Institute, Intel, Quanta Cloud Technology, Hong Kong ASTRI (Applied Science and Technology Research Institute) | Yan Yang | Jan-2020 | Started |
PoC#11: Intelligent Energy Management of DC | China Telecom Intel, AsiaInfo, Samsung, Huawei | Yu Zeng | March-2020 | Proposed |
The PoC description and presentation are available on the public ETSI ENI WiKi Pages (ENI Wiki). Progress reports were accepted from active Proof of Concepts. PoC#11 was agreed and sent to formal review.
A Drafting group was formed to work on a 2nd ENI whitepaper for the future of the ISG work, this drafting group comprised many member organisations: China Telecom, Huawei, Intel TIM, Samsung, Futurewei, AsiaInfo, Portugal Telecom, China Mobile, China Unicom CEA, Telefonica, Deutsche Telekom, NTT, University of Carlos III in Madrid and CAICT. Others are invited to join.
The ISG will now focus on the next version of the work.
Let’s remind our readers that the ISG is open to ETSI members and non-ETSI members alike. The different players in the value chain are welcome to join the ISG effort, contribute to the development of these key specifications and demonstrate Proofs of Concepts (PoCs) along with operators, vendors and research institutes from all over the world. To join, please contact: isgsupport@etsi.org
The next ENI plenary meeting (ENI#14) is planned to be hosted in Spain but may be hosted in a different place, depending on the evolution of the situation. There may be an associated conference collocated with the meeting.
The ETSI / ITU-T workshop on “ENI-Machine Learning in communication networks”
This workshop was also held online on 16 March 2020. It received presentations from ISG ENI, ISG SAI, ITU-T FG ML5G “Machine Learning 5th Generation” & ITU-T Study Group 13 Question 20. A guest speaker “Professor Stanczak: Technical University, Berlin” gave a presentation on “5G and beyond – where are we and what is next”. The workshop was open to the public. All presentations and related information are available from ETSI & from the ITU-T.
The workshop was especially useful to address synergies of the work. The use of a common Use Cases template was discussed in the afternoon. It was agreed that ETSI ISG ENI will draft a liaison to illustrate the use cases template used in GS ENI 001. This template includes a mapping to the ENI system architecture GS ENI 005, which can be generalised within the ITU-T.
It was questioned if this template could be used across all industries; ENI indicated that it was created to model the Telecommunications network, endpoints and storage clouds. It was yet to be seen if it could be industry independent or neutral.