NetOpt Day 2022

Optimisation and Algorithms for Networks

Sponsored by GdR RSD and the REST initiative from GdR RO

Date: November 10th 2022
Lieu: 292 rue St Martin, 75003, Paris. Salle des Conférences du Musée du Cnam.
Co-located with IEEE CloudNet 2022

Software-Defined Networks (SDN) make network control and management more flexible and intelligent. They represent a first step towards autonomous IP networks (“self-driving networks”) which consist in completely automated networks optimizing user experience or intents (also referred to as “Intent-Driven Networking (IDN)”). The predominance of software enable the extensive use of learning (ML/AI) and optimization (OR) techniques, previously reserved for offline planning. In this context, network controllers can benefit from significant computing power and a large amount of data to adapt the configuration of equipments at different time scales. Thus, controllers are able to analyze and predict traffic, predict network behavior and anticipate user experience, as well as to solve large-scale optimization problems (combinatorial, stochastic, robust) in an evolving environment. This transformation of network architectures is a good opportunity for the Operations Research (OR) community and the Network and Distributed Systems community to jointly provide answers to the many issues raised by this new paradigm.

Call for presentations:
This day aims at providing a forum for researchers to discuss and address the challenges related to the topics mentioned below. Researchers willing to present their work are invited to submit their presentation (30 min) before Septembre 30th 2022 (extended).

  • Application domains:
    • Traffic engineering
    • Network design and planning
    • Network slicing
    • Control plan architectures
    • Virtualization and cloud
    • Autonomous networks
    • Management of monitoring and security
    • Etc.
  • Mathematical and algorithmic tools:
    • Combinatorial and stochastic optimizaion
    • Online algorithms
    • Optimal control
    • Machine learning (statistical, reinforcement…)
    • Hybrid methods with learning (ML/IA) and operational research (OR), etc.
    • Etc.
Submissions consist of a title + short abstract proposal. They should be sent at The participation of PhD candidates is highly welcome. There won’t be proceedings. Presentations can be on already published or on progress work. Registration will be free but mandatory for logistics reasons.

  • Amal Benhamiche, Orange Labs
  • Frederic Giroire, CNRS
  • Jeremie Leguay, Huawei
  • Sebastien Martin, Huawei
  • Stefano Secci, CNAM