A load balancing algorithm for solving multi‐objective virtual network embedding

Abstract

Virtual network embedding (VNE) is one of the most basic problems in network virtualization. VNE is to deploy virtual network resources to substrate network resources to solve the problem of network rigidity, which is essentially a problem of resource optimization. It can be divided into two stages: node embedding and link embedding. Previous embedding schemes focused on the algorithm research of single objective, but in the process of VNE, many factors need to be considered, such as the benefits, acceptance rate, substrate resource load, and so on. Therefore, this article considered the multi‐objective VNE algorithm, proposed a load balancing algorithm for solving multi‐objective VNE, took the node load value as one of the fitness functions, and introduced global resource capacity to measure the embedding ability of the underlying nodes, initialized the population, improved the embedding accuracy, and used group search algorithm algorithm for iterative optimization. Simulation results showed that the algorithm can balance the load, effectively improve the acceptance rate of virtual network requests, and improve the revenue of user service providers.

Deployment and Reconfiguration for Balanced 5G Core Network Slices

“…For link mapping, a weighted shortest path method is utilized by assigning weights to links according to their resource utilization ratio. In [23], the authors propose a load balancing algorithm for solving a multi-objective VNE. The algorithm incorporate a VNE algorithm by measuring physical node load value for node mapping process.…”

“…Then, we try to remap these virtual links one-by-one by just remapping one-hop from the current congested physical node. After trying all virtual links in the MVL set, if the bandwidth deficit still exists, we go back to step 23 end if 30: end for the virtual links are tried for remapping and the deficit can not be solved, a rejection is returned that NS S+1 cannot be re-deployed.…”