Enhanced Anomaly Detection in Cloud Environments Using Deep Learning

End-to-end system communication has increased dramatically as a result of advancements in computer networking. Nevertheless, concerns about security have also been voiced. Consequently, it is still difficult to spot abnormalities in a complicated cloud environment. As a result, a deep Convolutional Neural Network (CNN) detection and classification model for near-real-time cloud network intrusions is proposed in this study. To choose the most appropriate characteristics to feed into the CNN model, the random forest model is also available and used. The experiments utilized CSE-CIC-IDS2018 datasets. With an error rate of just 2.93 percent and a testing accuracy of 97.07 percent, the proposed CNN model performed better than the competition. The accuracy, recall, and f1-score of the suggested model were 98.11, 96.93, and 97.52%, respectively, in these evaluations. These findings show great promise for real-time Industry 4.0 systems, since they are more accurate, precise, and capable of detecting network irregularities with the utmost precision.