基于粒子群优化的负载均衡的虚拟网络映射

Load-balancing Based on Particle Swarm Optimization in Virtual Network Mapping

虚拟网络映射忽略通信路径上中间节点的资源消耗,以底层物理网络资源消耗最小或路径最短为目标,导致最优路径上中间节点资源不足而出现瓶颈,影响整个底层网络的性能及后续虚拟网络请求成功率。该文针对这一问题考虑物理路径上中间节点资源消耗,以节点负载和链路负载同时达到均衡为目标,将路径跳数限制作为约束,建立虚拟网络映射的数学模型,并采用多目标负载均衡粒子群优化算法求解,记作LB-PSO。实验证明该算法有效地消除了资源瓶颈,为后续虚拟网络请求提供一个更为均衡的底层物理网络,从而提高了虚拟网络构建成功率、网络资源利用率以及基础设施提供商的收益。

Virtual networking mapping aims at minimal resource consumption or shortest path at substrate network,but ignores the resource demanded by the hidden hops,making bottlenecks due to the resource shortage at the hidden hops,and finally influences the performance of the entire substrate network and the request success rate of the consequential virtual network.Pointing at above problem and taking into account the resource demanded by the hidden hops,this paper aims at the simultaneous loading balance of the substrate node and the substrate link,mathematically formulate the virtual networking mapping problem constrained by hops,and solves it using multi-objective Load-Balancing Particle Swarm Optimization（LB-PSO） algorithm.This algorithm eliminates resource bottleneck efficiently,provides a more balanced substrate network for the request of the consequential virtual network request,thus improving the constructing success rate of virtual network,the availability of network resources and the profits of the infrastructure providers.