虚拟网络映射高效节能运输模型及算法

Transportation Model and Algorithms for Energy Efficient Virtual Network Embedding

网络虚拟化使得智能能量感知网络部署成为可能,已有研究忽略了节点映射能耗最优化.本文把节点映射能耗优化问题转化为生产地与销售地之间物资运输代价最优化问题,建立高效节能节点映射运输模型.根据最大元素法,提出了混合一阶段与两阶段映射算法,在链路映射的约束下找到节点分配最小能耗代价最优解;利用主动休眠策略,提出了基于运输模型的主动休眠虚拟网络映射节能算法;利用节点可重复映射技术,提出了基于运输模型的节点可重复映射算法,进一步提高了底层网络资源休眠数量.仿真结果验证了本文所提算法能够显著降低系统能耗,适合大规模高效节能虚拟网络映射.

Netw ork virtualization is an enabler for intelligent energy-aw are netw ork deployment. In the existing virtual netw ork embedding algorithms,the minimization of energy consumption for mapping virtual nodes is ignored. In this paper,w e create a transportation model for energy efficient virtual netw ork embedding,in w hich energy consumption minimization for mapping virtual nodes is turned into the transportation model. An algorithm based on the largest element is proposed to obtain the optimization solution w ith node and link resource constraints,w hich is also a hybrid one-and-tw o stage algorithm.Besides,on the basis of the transportation model,w e design an energy saving algorithm by use of the actively hibernating policy. If multiple virtual nodes in the same virtual netw ork can be mapped to the same substrate node,a largest-elementand-transportation-model-based algorithm is proposed. These algorithms increase the number of substrate netw ork hibernating resources. Simulation results show that the proposed algorithms can significantly reduce energy consumption. The theoretical analyses and the simulation results verify that our proposed algorithms suit energy efficient virtual netw ork embedding for large scale substrate netw ork.