摘要
因传统计算机网络内部丢包节点识别技术准确度不高,关键节点识别效率低且无法满足实际的应用需求,因此,提出基于CNN-LSTM(卷积神经网络)的计算机网络内部丢包节点识别技术方法。首先,在计算机网络内部丢包节点进行识别时,对计算机网络内部攻击模型进行分析;其次,建立网络信道模型,从而构建出节点交互成功的概率模型,提高交互成功的概率;最后,采用假设检验的异常节点识别。通过对比结果分析,随着恶意节点丢包率增加,导致选择性转发攻击路径的检测成功概率相应增加,证明了本文识别技术的检测功能具有实用性。
Due to the low accuracy of traditional packet loss node identification technology in computer networks,the low efficiency of key node identification,and the inability to meet practical application requirements,a method for identifying packet loss nodes in computer networks based on CNN-LSTM(Convolutional Neural Network) is proposed.Firstly,when identifying packet loss nodes within the computer network,the internal attack model of the computer network is analyzed;Secondly,a network channel model is established to construct a probability model for successful node interaction and improve the probability of successful interaction;Finally,abnormal node identification is performed using hypothesis testing.By comparing and analyzing the results,it can be concluded that as the packet loss rate of malicious nodes in the path increases,the probability of successful path detection increases.The method proposed in this paper is proved to be practical.
作者
吕航
LV Hang(Changchun Automotive Industry College,Changchun 130000,Jilin)
出处
《电脑与电信》
2023年第9期92-95,100,共5页
Computer & Telecommunication
关键词
计算机网络
内部
丢包节点
识别
computer network
internal
packet loss node
identification
