摘要
流量异常直接影响软件定义网络(SDN)的正常工作,为了解决当前SDN流量异常检测过程中存在的一些不足,以提升SDN流量异常检测精度,设计了基于深度学习的SDN流量异常检测系统。采用经验模态分解方法对SDN流量异常数据进行预处理,采用最小二乘支持向量机建立SDN流量异常检测模型,利用萤火虫算法实现最小二乘支持向量机参数优化选择,以获得更优的SDN流量异常检测结果。实验结果表明,本文系统的SDN流量异常值与实际值非常接近,大幅度减少了SDN流量异常检测误差,具有较高的实际应用价值。
Traffic anomaly directly affects the normal operation of software-defined network(SDN).In order to solve some deficiencies in the current SDN traffic anomaly detection process and improve the accuracy of SDN traffic anomaly detection,a deep learning-based SDN traffic anomaly detection system is designed.The empirical mode decomposition method is used to preprocess the abnormal SDN traffic data,the least squares support vector machine is used to establish the SDN traffic abnormality detection model,and the firefly algorithm is used to realize the optimal selection of the least squares support vector machine parameters to obtain better SDN traffic.Anomaly detection results.The experimental results show that the abnormal value of SDN traffic in the system in this paper is very close to the actual value,which greatly reduces the error of abnormal detection of SDN traffic,and has high practical application value.
作者
钟掖
龙玉江
赵威扬
张光益
ZHONG Ye;LONG Yujiang;ZHAO Weiyang;ZHANG Guangyi(Sichuan Normal University,Chengdu Sichuan Province 610066,China;Guizhou Power Grid Co.,Ltd.Information Center,550018,Guiyang,Guizhou Province 550000,China)
出处
《自动化与仪器仪表》
2022年第4期89-92,97,共5页
Automation & Instrumentation
关键词
软件定义网络
流量异常
萤火虫算法
最小二支持向量机
检测正确率
software-defined network
traffic anomaly
firefly algorithm
least-squares support vector machine
detection accuracy