A Virtual Network Embedding Algorithm Based On Hybrid Particle Swarm Optimization

Allocating the underlying physical substrate network resources for users reasonably is the target of virtual network embedding (VNE), which is a hot issue in virtual resource allocation field. In order to prevent the premature convergence and poor performance of local optimization during mapping procedure, in this paper, we combine DPSO, taboo-search technology and simulated annealing algorithms to solve premature convergence problem by using taboo list and annealing process, then propose a virtual network embedding algorithm based on hybrid particle swarm optimization. Simulation results show that our algorithm can improve the revenue to cost ratio and the acceptance ratio.