Virtual Network Embedding Based on Hybrid Adaptive Genetic Algorithm
Boyang Liu, Muqing Wu, Zou Haosen · 2019
Virtual network embedding (VNE) is one of the central parts of network virtualization, and its main objective is to maximize the revenue of substrate networks(SN). In the previous studies, heuristic algorithms are widely used to solve the embedding problems, but they are liable to trap in local optimum. In addition, they also have slow convergence rates which are caused by ignoring the intrinsic characteristics of the VNE. In this paper, on the first hand, we improve the cross operator of traditional genetic algorithm (GA) to an adaptive cross operator, in order to increase the convergence speed; on the other hand, a simulated annealing algorithm is used to replace the mutation operation, and the simulated annealing idea is adopted to modify the selection operation to avoid the local optimization. The simulation results show that the proposed algorithm has a significant improvement in the VNE acceptance rate and the revenue-to-cost ratio.