基于粒子群优化算法的数据中心网络流量调度策略

A Traffic Scheduling Strategy Based on Particle Swarm Optimization in Data Center Network

为解决当前数据中心网络存在链路负载不均衡及带宽资源浪费问题,提出了一种基于粒子群优化算法的流量调度策略。该策略结合软件定义网络控制器可获取全局网络拓扑信息的特性,依据当前链路带宽资源状况及网络流量的带宽需求建立目标函数。首先,根据流的源地址和目的地址找出最短路径集,通过定义粒子聚合度判断算法是否有陷入局部最优的趋势;然后,结合约束条件与目标函数,利用优化的粒子群算法从最短路径集中找出网络流量的最佳调度路径。实验结果表明,相比于其他算法,该算法有效地提高了网络平均吞吐量,获取了较低的丢包率,从而减轻了带宽资源的浪费,更好地实现了网络的负载均衡。

To solve the problems of unbalanced link load and waste of bandwidth resource in the current data center network,a traffic scheduling strategy based on particle swarm optimization algorithm(TSM_PSO)is proposed.The strategy combines the characteristic of the software defined network(SDN)controller that can obtain global network topology information,and establishes the objective function based on the current status of link bandwidth resources and the bandwidth requirements of the flow.First of all,it searches for the shortest path set based on the source and destination address of the flow,and determines whether the algorithm has local optimal trend by defining the aggregation of the particles.Then,combining the constraints with objective function,it makes use of TSM_PSO to find out the best scheduling path of flow from the shortest path set.The experiment results show that compared with other algorithms,the proposed algorithm effectively improves the average network throughput,and reduces the packet loss rate of flows,thus reducing the waste of bandwidth resources and achieving network load balancing.