为了解决软件定义网络(Software Defined Network,SDN)网络流量测量的节点选择中,受环境影响因素导致选择节点的效率低下和估计精度不够的问题.研究以蚁群优化的测量节点选择方法和小流推测的异常检测机制为基础,在蚁群优化算法的基础上加入领域搜索算法进行改进;并且提出以小流推测为基础的网络异常检测机制,对多种网络安全异常进行识别.实验结果显示,改进的蚁群优化算法(Ant Colony Optimization,ACO)算法准确性由0.504提高到1.000;收敛性由0.483提高到0.721;单位时间开销由0.905控制降低到了 0.105.数据表明优化后的ACO算法在SDN网络中流量测量的精确度得到了提高.以小流推测为基础的网络异常及检测方法在网络安全实验中表现出了优良的识别异常的能力,可以广泛应用在数据安全保障方面.
Application Research of Traffic Measurement in SDN Network Security Diagnosis
In order to solve the problem of low efficiency and insufficient estimation accuracy in node selection for software de-fined network(SDN)traffic measurement due to environmental factors.Based on the measurement node selection method of ant colo-ny optimization and the anomaly detection mechanism of small flow speculation,a domain search algorithm is added to improve the ant colony optimization algorithm;And propose a network anomaly detection mechanism based on small flow speculation to identify various network security anomalies.The experimental results show that the accuracy of the improved Ant Colony Optimization(ACO)algo-rithm has been improved from 0.504 to 1.000;The convergence has been improved from 0.483 to 0.721;The unit time cost has been reduced from 0.905 to 0.105.The data shows that the optimized ACO algorithm has improved the accuracy of traffic measurement in SDN networks.The network anomaly and detection method based on small flow speculation has shown excellent ability to identify a-nomalies in network security experiments,and can be widely applied in data security assurance.
ant colonyoptimization algorithmnetwork security new candidate set definitionmedical data
万方数据
维普
