基于蚁群优化算法的SDN路由策略

SDN Routing Strategy Based on Ant Colony Optimization Algorithm

针对蚁群算法在软件定义网络路由选择中的全局搜索能力弱、收敛速度慢的问题,提出一种基于蚁群优化算法的路由策略.根据网络规模设定参数,将信息素浓度重要程度和挥发系数由静态参数改进为动态参数,弱化算法迭代前期的信息素浓度重要程度以提升算法前期的全局搜索能力,增强算法迭代后期信息素浓度重要程度以加快算法后期的收敛速度;对挥发系数采用逐步减小的动态参数使算法避免陷入局部最优解;进一步加快算法后期的收敛速度,使网络获取更佳性能.在Mininet平台上进行仿真实验评估该算法性能,实验表明该算法前期在选择路由时的全局搜索能力增强,后期收敛速度明显加快.实验通过将基于蚁群优化算法的SDN路由策略与基于最短路径路由算法、等价多路径路由算法路由策略对比,链路利用率分别提升9.9%和17.1%,具有平均吞吐量大、链路利用率高的优点.

Aiming at the problems of weak global search ability and slow convergence speed of ant colony algorithm in software-defined network routing,a routing strategy based on ant colony optimization algorithm is proposed.We set parameters according to network scale,improve pheromone concentration importance and volatilization coefficient from static parameters to dynamic parameters,weaken pheromone concentration importance at the early stage of algorithm iteration to improve global search ability at the early stage of algorithm iteration,and enhance pheromone concentration importance at the late stage of algorithm iteration to accelerate convergence speed at the later stage of algorithm iteration.The algorithm avoids falling into local optimal solution by using dynamic parameters which gradually decrease the volatilization coefficient.We accelerate the convergence speed in the later stage of the algorithm,and make the network obtain better performance.The performance of the algorithm is evaluated by simulation experiments on Mininet platform.The experiments show that the global search ability of the algorithm is enhanced in the early stage of routing selection,and the convergence speed is obviously accelerated in the later stage.By comparing SDN routing strategy based on ant colony optimization algorithm with routing strategy based on shortest path routing algorithm and equivalent multipath routing algorithm,the link utilization rate is increased by 9.9%and 17.1%respectively,which has the advantages of large average throughput and high link utilization rate.