针对虚拟化网络环境中的资源分配问题,通过深度优先搜索遍历虚拟网络,构造相邻的虚拟节点队列.根据网络的拓扑结构以及节点和链路的资源状态,自适应地扩展物理网络拓扑结构,协调地将相邻的虚拟节点和其邻接链路映射到负载强度较低的邻接物理节点和物理链路上.仿真结果表明,AAG—VNM算法有效地降低了虚拟网络映射的资源开销,提高了物理网络资源利用率和虚拟网络请求接受率.
Aiming at the resource allocation problem in the virtualization network environment, a neighbor queue of virtual nodes is constructed by using the depth-first search traversing. The substrate network is adaptively augmen- ted based on network structure, nodes and links' resource state. Then neighbor virtual nodes and their neighbor links are mapped into the neighbor substrate nodes and substrate links of low load stress in a coordinated way. Simulation experiments show that the proposed algorithm decreases resource cost of embedding virtual networks while increases substrate network resource utilization ratio and acceptance ratio of virtual network requests compared with the traditional method.