Performance Modeling and Joint Resource Allocation Algorithms for Online Virtual Network Embedding

Network Virtualization (NV) has been proposed as an enabling technology of a key value-added service for service providers. While there are a very large number of publications that have proposed various resource allocation algorithms for NV, no effort has been made to estimate the performance of vir...

Full description

Saved in:
Bibliographic Details
Published in:IEEE eTransactions on network and service management 2024-02, Vol.21 (1), p.1048-1066
Main Authors: Lu, Qiao, Huang, Changcheng
Format: Article
Language:English
Subjects:
Algorithms
Analytical models
analytical performance estimation
Embedding
Genetic algorithms
Heuristic algorithms
Internet resources
loss network model
Mathematical analysis
Mathematical models
Network topology
Network virtualization
Recursive functions
Resource allocation
Resource management
Servers
Substrates
Topology
virtual network embedding
Virtual networks
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Network Virtualization (NV) has been proposed as an enabling technology of a key value-added service for service providers. While there are a very large number of publications that have proposed various resource allocation algorithms for NV, no effort has been made to estimate the performance of virtual network embedding (VNE) algorithms based on analytical models. In this paper, to assess the blocking probability of virtual link mapping, we propose a novel loss network model with Dynamic Routing And Random Topology (DRART). Moreover, by combining with some existing models, we can estimate the blocking probability for VNE through a creative recursive process. Our model can provide a benchmark for various VNE algorithms. To fill the performance gap between existing resource allocation algorithms and our analytical model, we also propose a distributed Genetic Algorithm (GA) based resource allocation approach that can jointly allocate node and link resources. Our simulation results show that our resource allocation approach can achieve the performance as predicted by our analytical model while meeting stringent online resource allocation requirements.
ISSN:1932-4537
1932-4537
DOI:10.1109/TNSM.2023.3308065