On Combining Reinforcement Learning and Monte Carlo for Dynamic Virtual Network Embedding - Archive ouverte HAL Access content directly
Poster Communications Psycho-Social-Medicine Year : 2021

On Combining Reinforcement Learning and Monte Carlo for Dynamic Virtual Network Embedding

(1) , (1) , (2)
1
2

Abstract

Network slicing is one of the key building blocks in the evolution towards "zero touch networks". Indeed, this will allow 5G and beyond 5G networks to deploy services dynamically, on the same substrate network, regardless of their constraints. In this demo, we introduced a platform for dynamic virtual network embedding, a problem class known to be NP-hard. The proposed solution is based on a combination of a deep reinforcement learning strategy and a Monte Carlo (MC) approach. The idea here is to learn to generate, using a Deep Q-Network (DQN), a distribution of the placement solution, on which a MC-based search technique is applied. This makes the agent's exploration of the solution space more efficient.
Fichier principal
Vignette du fichier
1570760840 final.pdf (204.62 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03510747 , version 1 (08-01-2022)

Identifiers

  • HAL Id : hal-03510747 , version 1

Cite

Ghina Dandachi, Yassine Hadjadj-Aoul, Abdelkader Outtagarts. On Combining Reinforcement Learning and Monte Carlo for Dynamic Virtual Network Embedding. LCN 2021 - 46th IEEE Conference on Local Computer Networks, Oct 2021, Edmonton, Canada. IEEE, pp.1-3, 2021. ⟨hal-03510747⟩
57 View
73 Download

Share

Gmail Facebook Twitter LinkedIn More