SDN中基于遗传蚁群优化的测量节点选择方案设计

Measurement Node Selection Scheme Design Based on Genetic Ant Colony Optimization in SDN

SDN中的测量节点选择问题其本质为最小顶点覆盖模型的求解,然而流量路由信息的保护造成基于流的动态测量节点选择方案失效,只能转向静态测量节点选择,目前的静态测量节点选择算法存在测量精度较低、运行时间长等问题。本文针对SDN中测量节点选择算法性能受限问题,设计了一个基于遗传和蚁群算法的测量节点选择方案,该方案所设计的G-ACO算法将遗传算法和蚁群算法进行动态融合,并将其应用于解决最小顶点覆盖问题,有效提升最小顶点覆盖模型求解速度。最后,在SDN仿真网络环境下以4个不同数量级的网络拓扑进行节点选择方案验证,与其他算法进行对比分析,G-ACO遗传蚁群优化算法具有更好的寻优能力、算法稳定性。

The essence of measurement node selection problem in SDN is solving the minimum vertex cover model.However,protection for the traffic routing information leads to the failure of flow-based dynamic measurement node selection scheme,having to choose the static measurement node selection.Available static measurement node selection algorithms have the problems of low measurement accuracy and long running time.To address the problem of poor performance of measurement node selection algo-rithm in SDN,in this paper a measurement node selection scheme based on genetic and ant colony algo-rithms is proposed,where a G-ACO algorithm is designed to dynamically integrate the genetic algorithm and the ant colony algorithms to solve the minimum vertex cover model and to effectively improve the solving speed.Finally,the proposed scheme is verified in four network topologies with different orders of magnitude in SDN simulation network environment.The G-ACO genetic ant colony optimization algo-rithm outperforms other algorithms,and has higher optimization searching ability and algorithm stabili-ty.