基于滑动区域的粒子群虚拟网节能映射算法

针对传统虚拟网节能映射中存在的节点映射分散、链路映射跳数多等问题，利用虚拟网请求的最小生成树拓扑将节点和链路同时映射，该文提出了基于滑动区域的粒子群虚拟网节能映射算法(EVNE_SRPS)。当一个虚拟网请求到达时，生成其最小生成树拓扑，根节点为路径和最短的节点；在底层网络随机选取多个区域作为粒子对象，并在区域中心映射虚拟网请求的最小生成树拓扑；计算粒子的适应度，求出群体和个体最优解，并在最优解的指导下确定滑动方向、更新区域位置，经过迭代后得到虚拟网的映射方案。实验结果表明，与现有算法相比，该算法降低了网络能耗，提高了运营商的收益成本比。

Energy-saving Virtual Network Embedding Algorithm Based on Sliding Region Particle Swarm

Considering the problem of scattered node mapping and more hops of link mapping in the traditional virtual network energy-saving embedding, the node and link are mapped simultaneously by using the minimum spanning tree topology of the virtual network request, and Energy-saving Virtual Network Embedding algorithm based on Sliding Region Particle Swarm (EVNE_SRPS) is proposed. When a virtual network request arrives, the minimum spanning tree topology is generated, the root node is the node with the shortest path length; Multiple regions are randomly selected as the particle object in the substrate network, and the minimum spanning tree topology of the virtual network request is mapped in the regional center; The fitness of the particles is calculated. The optimal solution of the group and the individual is finded, and the sliding direction and the location of the update region under the guidance of the optimal solution are determined. After the iteration, the mapping scheme of the virtual network is obtained. The experimental results show that compared with the existing algorithms, the network energy consumption is reduced, and the internet service providers revenue to cost ratio is improved.