针对虚拟网络映射中的能耗问题,根据虚拟网络重构特征以及节点和链路的能耗特性,建立虚拟网络重构优化模型。通过设置底层物理网络的资源利用率阈值,周期性地将资源利用率大于高阈值和小于低阈值物理节点和物理链路上映射的虚拟节点和虚拟链路迁移到能耗增幅较小的物理节点和物理链路上,并采用节点、链路休眠和唤醒机制,动态地调整网络中活动物理节点和物理链路数量。模拟结果表明:EE-VNR算法有效地均衡了底层物理网络中节点和链路的负载水平,提高了虚拟网络请求接受率,大大地降低了虚拟网络映射的系统能耗。
Aiming at the problem of energy consumption in the virtual network mapping,according to the characteristics of the virtual network reconfiguration and energy consumption of substrate nodes and links,a reconfiguration optimization model of mapping virtual networks is proposed. By setting resource utilization threshold in the substrate network,virtual nodes that are already mapped onto substrate nodes,which resource utilization is greater than the high threshold or less than low threshold,are remapped onto the substrate nodes consuming smaller energy periodically,the same operation as the virtual links. By using the method of hibernating and waking up substrate nodes and links,the number of active substrate nodes and substrate links in the substrate network are adjusted dynamically. Simulation results show that EEVNR algorithm balances the substrate nodes and links load effectively,improves the acceptance ratio of virtual network requests and reduces the system energy consumption of mapping virtual networks evidently.