数字图书馆公用网络信息传输通道恶意节点检测方法

Malicious node detection method for public network information transmission channel of digital library

以确保数字图书馆用户信息的安全性为目的,研究了数字图书馆公用网络信息传输通道恶意节点检测方法。通过拓扑结构分析网络场景,以有效发送率、转发率、入度与传输时延均值描述节点特征属性,再根据节点特征属性构建恶意节点攻击模型。根据恶意节点攻击特征全方位检测网络节点的运行状态并构建观测序列,然后训练隐半马尔科夫模型,通过确定观测序列对于隐半马尔科夫模型的熵值判断节点是否为恶意节点。实验结果表明:该方法能够有效描述实验对象内的恶意节点攻击行为,并准确检测恶意节点。

In order to ensure the security of digital library user information,this paper studies the malicious node detection method of digital library public network information transmission channel.The network scene is analyzed by topology structure,and the characteristic attributes of nodes are described by effective transmission rate,forwarding rate,inbound rate and average transmission delay,and then the malicious node attack model is constructed according to the characteristic attributes of nodes.According to the attack characteristics of malicious nodes,the running state of network nodes is comprehensively detected and the observation sequence is constructed.Then the hidden semi-Markov model is trained,and whether the nodes are malicious nodes is determined by determining the entropy value of the hidden semi-Markov model.Experiment results show that this method can describe the malicious attack behavior of nodes in the experimental object effectively and detect malicious nodes accurately.