摘要
为提高计算机网络的使用安全,引进机器学习算法改进计算机网络节点漏洞检测方法。将节点流量熵值作为识别计算机网络节点异常流量的关键依据,根据窗口大小计算节点流量熵值;提取计算机网络环境中软件定义网络(Software Defined Network,SDN)控制器的运行参数,识别计算机网络节点异常流量;引进机器学习(Machine Learning,ML)算法,将流通节点的数据输入到机器学习网络层中,训练计算机网络节点数据;利用自编码技术将高维非线性数据转换为低维数据,在网络节点之间建立双向映射关系,以自底向上的方式检测节点漏洞。进行对比实验,结果表明,设计的方法可以精准检测网络环境中的异常数据与漏洞节点。
To improve the security of computer network usage,machine learning algorithms are introduced to improve computer network node vulnerability detection methods.The node traffic entropy value is used as a key basis for identifying abnormal traffic of computer network nodes,and the node traffic entropy value is calculated based on the window size;the operating parameters of the Software Defined Network(SDN)controller in the computer network environment are extracted to identify abnormal traffic of computer network nodes;the Machine Learning(ML)algorithm,input data from circulation nodes into the machine learning network layer to train computer network node data;use self-coding techniques to convert high-dimensional nonlinear data into low-dimensional data,and establish bidirectional mapping relationships between network nodes to detect node vulnerabilities in a bottom-up manner.Comparative experiments are conducted,and the results show that the designed method can accurately detect abnormal data and vulnerable nodes in the network environment.
作者
邱增烇
QIU Zengzheng(Shanghai Hengnan Intelligent Technology Co.,Ltd.,Shanghai 201203,China)
出处
《信息与电脑》
2022年第21期111-113,共3页
Information & Computer
