Abstract : Identifying cyber-attacks in cloud infrastructures is essential for protecting the cloud environment from cyber-attacks. It is difficult to detect cyber-attacks in cloud infrastructures due to the complex and distributed nature of cloud infrastructures. In addition, various attacks that happen dynamically or randomly, online attacks, adversarial attacks, and data leakages increase the difficulty and complexity of cyber-attack detection. The work has proposed highly secured semi supervised anomaly detection in a cloud environment using the IRLS-cGAN detection technique. To enhance the capability of the proposed framework, the work has explored Data preprocessing, data extraction, data selection, and finally, attack classification. The approach arrests the probability of attack caused due to unnecessary data by providing a proper structuration with the help of data preprocessing that includes redundant data removal, handling categorical features, data scaling, and handling imbalanced data. In order to bring an optimal relation between the features and the outcome, the approach has extracted the informative data using GRA and selected the relevant data using the MRSO technique. Thereafter, the behavior of various environments based on clouds, VM, networking, and attacks are grouped into a cluster using the CSI-HAC technique, which is then trained to IRLS-cGAN for detecting the attacks. The results obtained demonstrate that the proposed cloud-based anomaly detection model is superior in comparison to the other state-of-the-art models (used for network anomaly detection) in terms of accuracy, detection rate, false-positive rate, and false-negative rate.

Keywords: Cloud Computing, Virtual Machine (VM), cyber security, anomaly detection, Multicollinearity, Grey Relation Analysis (GRA), Mutated Rat Swarm Optimization (MRSO), Cauchy Schwartz Inequality-Based Hierarchical Agglomerative Clustering ( CSI-HAC), and Iteratively Reweighted Least-Squares-Based Conditional Generative Adversarial Network (IRLS-cGAN).

| DOI: 10.17148/IARJSET.2021.8846