A PSO-Based Virtual Network Mapping Algorithm with Crossover Operator

摘要

Virtual network mapping (VNM) is a crucial technology for network virtualization to allocate network resource. One of the major challenges for virtual network mapping is the efficient allocation of substrate resources to the virtual networks. In order to further improve the efficiency of the previous algorithms in large scale network, the crossover operator is introduced into discrete particle swarm optimization algorithm, and a hybrid intelligent algorithm is designed. The algorithm can solve the problem that the traditional particle swarm algorithm is easy to fall into local optimal point and is difficult to reach the global optimal. Experimental results show that the algorithm consumes the lowest cost in the case of mapping the same virtual network requests than existing ones, and has higher revenue/costs ratio.