面向动态环境的虚拟光网络资源优化算法

Resource Optimization Algorithm for Virtual Optical Networks under Dynamic Environment

为解决采用单个虚拟网特征重配置导致虚拟网接受率低的问题,基于资源竞争特征重构底层网络模型,并基于此构建虚拟网映射线性规划模型。设计了无划分的随机舍入优化算法和有划分的随机舍入优化算法,为每个虚拟网分配全局最优的底层网络资源。与现有算法相比,该算法下虚拟网请求的平均接受率提升11%,节点资源平均利用率提升61%,链路资源平均利用率提升27%。

In order to solve the problem of low acceptance rate of virtual network caused by reconfiguration of single feature of virtual network,the substrate network model is reconstructed based on resource competition characteristics,and a linear programming model of virtual network mapping is constructed.A random rounding optimization algorithm without partition and a random rounding optimization algorithm with partition are designed to allocate globally optimal substrate network resources for each virtual network.Compared with the existing algorithm,the average acceptance rate of virtual network requests under the proposed algorithm is increased by about 11%,the average utilization rate of node resources is increased by about 61%,and the average utilization rate of link resources is increased by about 27%.