There is a newer version of the record available.

Published August 7, 2022 | Version v1
Project deliverable Open

DAEMON Deliverable 2.2: Initial DAEMON Network Intelligence framework and toolsets

  • 1. Fundación IMDEA Networks
  • 2. i2CAT
  • 3. IMEC
  • 4. Nokia Bell Labs
  • 5. NEC Laboratories Europe GmbH
  • 6. OTE
  • 7. Telefónica Investigación y Desarrollo SA
  • 8. Universidad Carlos III de Madrid
  • 9. Universidad de Málaga
  • 10. WINGS
  • 11. Zettascale

Description

This is the second public deliverable of WP2 of the DAEMON project. It builds upon the material of the previous deliverable of WP2, i.e., D2.1, and on activities and results achieved during the first iteration of the project in WP3, WP4 and WP5. As a result, the document describes the following content.

First, it provides significant updates to the functional and non-functional requirements of the eight NI-assisted functionalities tackled by DAEMON. These updates are mainly driven by the feedback provided by the implementation and evaluation activities of the project to date, which unveiled the need for adjustments to the original requirements or the addition of completely new ones.

Second, it defines novel and equivalent functional and non-functional requirements for a NI plane, which the project proposes to complement existing network planes (e.g., user, control, management), and which lay the foundations to the modelling of the NI architectural framework targeted by the project.

Third, based on the aforementioned requirements, it draft a preliminary yet complete design of the DAEMON’s NI plane, detailing its organization, its interaction with complementary blocks (e.g., traditional networking planes and machine learning operations), and its functioning in terms of NI instance coordination.

Fourth, set forth a first set of project’s guidelines for the pragmatic design of NI. These span along two directions. First, a NI design tailored to the needs of B5G network management, orchestration and control. Second, a NI design that does not stretch the adoption of complex data-hungry black-box models for all possible any NI use cases; instead, it considers using more traditional, simpler, or interpretable models across the functionalities addresses in the project.

Fifth, it summarizes the current status of a continuous review of the literature on NI and machine learning for mobile network management. This is a live effort that aims at keeping the project activities up to date with the current state of the art in scientific research, as well as to ultimately produce a complete survey and taxonomy of solutions for NI design and integration.

As a result, the content of this document will be the basis on which the rest of the second iteration of the project activities will build upon. We expect that the updated design implementation of NI-assisted functionalities throughout the second iteration will be driven by the requirements set forth in this deliverable. Similarly, the alignment of the NI algorithm for orchestration via the NI plane will be carried out respect to the representation models presented in this deliverable, Finally, all evaluations will be aimed at verifying that the proposed solutions meet the project KPIs in terms of performance, but also the functional and non-functional requirements and NI design guidelines listed in this deliverable.

Files

ict52-daemon-deliverable-2.2.pdf

Files (3.8 MB)

Name Size Download all
md5:2e73abd32e6798966c6edbf72bdd2324
3.8 MB Preview Download

Additional details

Funding

European Commission
DAEMON - Network intelligence for aDAptive and sElf-Learning MObile Networks 101017109