基于深度学习的SDN环境下异常流量检测方法

Abnormal Traffic Detection Method in SDN Based on Deep Learning

针对传统的异常检测方法在部署在SDN网络中,存在算法复杂、计算开销大并且产生额外流量的问题,提出一种基于深度学习的轻量型异常流量检测方法。通过分析流量数据特征重要度,构建异常检测数据,利用循环神经网络提取检测数据关联性信息,并利用轻量型分类函数实现对异常流量的识别。实验结果表明,所提方法较传统的异常流量检测方法在精确率、召回率等指标上有明显优势,且具有模型结构简单,部署方便,对SDN控制器性能影响小的特点。

Aiming at the problems that traditional anomaly detection methods are complicated in algorithm,high in calculation cost and generate extra traffic when deployed in SDN network,a lightweight anomaly traffic detection method based on deep learning is proposed.By analyzing the importance of traffic data features,detection data is constructed,correlation information of detection data is extracted by using circular neural network,and anomaly traffic is detected by using lightweight classification function.The experimental results show that the proposed method has obvious advantages over the traditional detection methods in terms of accuracy,recall and detection time,and has the characteristics of simple deployment and little impact on the performance of SDN controller.