摘要
Abstract
In cloud computing environments,where large-scale,high-dimensional data streams need to be processed,detection methods frequently fail to meet real-time requirements,leading to slow algorithm convergence and low accuracy.A network traffic anomaly detection model is proposed for cloud computing environments.Firstly,in a cloud computing environment,establish a network traffic communication model and update node parameters to initialize the traffic data process.Subsequently,principal component analysis is used to reduce the dimensionality and noise of high-dimensional traffic features to extract the principal components of abnormal features.Lastly,to construct a deep learning model integrating convolutional neural networks and gated recurrent units to achieve collaborative learning and anomaly classification of spatial and temporal features in network traffic.The comparative experimental results demonstrate that the proposed model achieves an F1 value of 98.0%on the test set,which is 7.8%and 3.5%higher than the improved dual depth Q-network and stacked convolutional attention model,respectively.Furthermore,its convergence speed is faster,and it can effectively achieve real-time accurate detection of network traffic and abnormal partition localization.关键词
云计算环境/流量特征学习/卷积神经网络/主成分提取/异常检测模型/网络流量Key words
cloud computing environment/traffic feature learning/Convolutional Neural Network/principal component extraction/anomaly detection model/network flow分类
航空航天
