A Heuristically Assisted Deep Reinforcement Learning Approach for Network Slice Placement

Network Slice placement with the problem of allocation of resources from a virtualized substrate network is an optimization problem which can be formulated as a multi-objective Integer Linear Programming (ILP) problem. However, to cope with the complexity of such a continuous task and seeking for op...

Full description

Saved in:
Bibliographic Details
Published in:IEEE eTransactions on network and service management Vol. 19; no. 4; pp. 4794 - 4806
Main Authors: Alves Esteves, Jose Jurandir, Boubendir, Amina, Guillemin, Fabrice, Sens, Pierre
Format: Journal Article
Language:English
Published: New York IEEE 01.12.2022
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects:
Algorithms
automation
Bandwidth
Computer Science
Convergence
Deep learning
deep reinforcement learning
Feature extraction
Integer programming
large scale
Linear programming
Machine learning
Network slicing
Networking and Internet Architecture
Optimization
Placement
Random access memory
Reinforcement learning
Substrates
Training
Automation
Deep Reinforcement Learning
Placement
Large Scale
Optimization
Network Slicing
ISSN:1932-4537, 1932-4537
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Network Slice placement with the problem of allocation of resources from a virtualized substrate network is an optimization problem which can be formulated as a multi-objective Integer Linear Programming (ILP) problem. However, to cope with the complexity of such a continuous task and seeking for optimality and automation, the use of Machine Learning (ML) techniques appear as a promising approach. We introduce a hybrid placement solution based on Deep Reinforcement Learning (DRL) and a dedicated optimization heuristic based on the "Power of Two Choices" principle. The DRL algorithm uses the so-called Asynchronous Advantage Actor Critic (A3C) algorithm for fast learning, and Graph Convolutional Networks (GCN) to automate feature extraction from the physical substrate network. The proposed Heuristically-Assisted DRL (HA-DRL) allows for the acceleration of the learning process and substantial gain in resource usage when compared against other state-of-the-art approaches, as evidenced by evaluation results.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1932-4537
1932-4537
DOI:10.1109/TNSM.2021.3132103