基于集成神经网络的通信网络入侵节点快速检测  

Rapid Detection of Intrusion Node in Communication Network Based on Integrated Neural Network

作　　者：李英楠 李璠 LI Yingnan;LI Fan(Gansu Power Investment Changle Power Generation Co.,Ltd.,Jiuquan 736100,China)

机构地区：[1]甘肃电投常乐发电有限责任公司,甘肃酒泉736100

出　　处：《通信电源技术》2025年第4期168-170,共3页Telecom Power Technology

摘　　要：为实现对通信网络中异常节点的快速检测,提升网络通信的质量,基于集成神经网络对其入侵节点快速检测方法展开全面的设计与研究。利用Python中的socket模块,探测通信网络传输层的用户数据报协议(User Datagram Protocol,UDP)端口,实现通信网络节点的扫描;将采样的数据录入集成神经网络,扫描节点的流量数据,对其进行训练并提取异常数据;计算节点异常程度得分,进行入侵节点检测与故障点快速定位。对比实验结果表明,设计的方法可以准确检测通信网络节点入侵。To achieve rapid detection of abnormal nodes in communication networks and improve the quality of network communication,a comprehensive design and research on intrusion node rapid detection method based on integrated neural network is carried out.Using the socket module in Python,probe the User Datagram Protocol(UDP)ports of the communication network transport layer to scan communication network nodes;Input the sampled data into an integrated neural network,scan the traffic data of nodes,train them,and extract abnormal data;Calculate the score of node anomaly degree for intrusion node detection and rapid location of fault points.The comparative experimental results show that the designed method can achieve accurate detection of intrusion into communication network nodes.

关 键 词：集成神经网络 异常数据 节点训练 快速检测 入侵节点 通信网络 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术] TP183[自动化与计算机技术—计算机科学与技术]