Automatic Virtual Network Embedding Based on Deep Reinforcement Learning

The performance of virtual network embedding determines the effectiveness and efficiency of a virtualized network, making it a critical part of the network virtualization technology. However, most existing algorithms fail to provide automatic embedding solutions in an acceptable running time. In thi...

Full description

Saved in:
Bibliographic Details
Published in2019 IEEE 21st International Conference on High Performance Computing and Communications; IEEE 17th International Conference on Smart City; IEEE 5th International Conference on Data Science and Systems (HPCC/SmartCity/DSS) pp. 625 - 631
Main Authors Yan, Zhongxia, Ge, Jingguo, Wu, Yulei, Zheng, Hongbo, Li, Liangxiong, Li, Tong
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.08.2019
Subjects
Bandwidth
Conferences
Feature extraction
Graph Convolutional Network
Kernel
Network topology
Network Virtualization
Reinforcement Learning
Substrates
Training
Virtual Network Embedding
Online AccessGet full text

Cover

More Information
Summary:The performance of virtual network embedding determines the effectiveness and efficiency of a virtualized network, making it a critical part of the network virtualization technology. However, most existing algorithms fail to provide automatic embedding solutions in an acceptable running time. In this paper, we combine reinforcement learning with a novel neural network structure and propose a new virtual network embedding algorithm. The proposed algorithm can learn to embed virtual networks automatically. Extensive simulation results show that our algorithm achieves the best performance on most metrics compared with the existing typical and stateof-the-art solutions.
DOI:10.1109/HPCC/SmartCity/DSS.2019.00095