An intelligent deep learning method for network intrusion detection in a cloud computing environment

Cloud computing has evolved rapidly through advanced technologies and applications; however, security remains a major concern. Security frameworks, such as access control, firewalls, confidentiality, and authentication, are essential for ensuring cloud security. In addition, the huge scale, high dimensionality, and great redundancy of network activity in cloud environments make it difficult to design effective cloud security solutions. Intrusion detection systems (IDSs) are critical for protecting cloud environments from diverse security threats and criminal activities. To address this challenge, it presents a deep learning (DL)- based framework that automatically learns robust feature representations for accurate and efficient intrusion detection. Here, proposed an effective unsupervised feature extraction method, that is, stacked contractive autoencoder (SCAE). The proposed approach uses a stacked contractive autoencoder (SCAE) to develop optimized low-dimensional representations from raw network traffic data. These extracted features are then classified using a hybrid extreme gradient boosting and support vector machine (EXGBoost-SVM) framework to build an effective cloud intrusion detection system (IDS). The proposed SCAE-EXGBSVM framework effectively leverages the complementary strengths of both approaches, thereby reducing computational complexity while improving detection performance. In terms of detection accuracy, robustness, and overall classification performance, the suggested SCAE-EXGBSVM model consistently outperformed five baseline methods, according to simulation evaluations conducted on two popular benchmark datasets, CICIoT2023 and UNSW-NB15.