信息安全研究2025，Vol.11Issue(3)：241-248,8.DOI:10.12379/j.issn.2096-1057.2025.03.06
一种融合时空特征的物联网入侵检测方法
An Intrusion Detection Method for Internet of Things by Fusing Spatio-temporal Features
摘要
Abstract
Aiming at the problems of insufficient attack samples and more categories in unbalanced IoT traffic datasets reducing the classification accuracy and generalization ability of the detection model,an intrusion detection method for the Internet of things by fusing spatio-temporal features(BGAREU)is proposed.The data were first normalized and the SMOTEENN method was used to improve the data distribution of the training samples;then temporal features and global information were extracted by Bi-directional gated recurrent unit(BiGRU)and multi-head attention,and combined ResNext network and U-Net network to construct a multi-scale spatial feature extraction network,and then incorporate efficient channel attention(ECA-Net)into the residual units to enhance the local characterization capability;finally,the fused features are fed into the Softmax classifier for multi-classification.Experiments show that the proposed model has more than 2％improvement in all the metrics compared with other models on IoT traffic datasets UNSW-NB15,NSL-KDD,and WSN-DS.In addition,this paper verifies that the ECA-Net has stronger characterization ability by comparing multiple attention mechanisms,and explores the effect of different numbers of attention heads in multi-head attention on the model performance.关键词
入侵检测/双向门控循环单元/多头注意力/多尺度特征提取/高效通道注意力Key words
intrusion detection/bidirectional gated recurrent unit/multi-head attention/multi-scale feature extraction/ECA-Net分类
计算机与自动化引用本文复制引用
翁铜铜,矫桂娥,张文俊..一种融合时空特征的物联网入侵检测方法[J].信息安全研究,2025,11(3):241-248,8.基金项目
国家自然科学基金面上项目(42376194) （42376194）
上海市科技创新行动计划项目(19511104502) （19511104502）
上海科学技术委员会科普项目(19DZ22048) （19DZ22048）
