Dynamic PHAD / AHAD Analysis for Network Intrusion Detection and Prevention System for Cloud Environment
Anitha Thangasamy, Bose Sundan, Logeswari Govindaraj · 2021 4th International Conference on Computing and Communications Technologies (ICCCT) · 2021
Nowadays, cloud computing plays an important role in organizations that offer a lot of resources and computing facilities on the internet. Due to its desirable features, a huge number of users are engaged in using cloud systems. As a result, traffic information is attacked both internally and externally in cloud systems. Hence, it is important to formulate an appropriate Intrusion Detection System with high accuracy. Earlier, normal and malicious attacks were detected during traffic information by IDS-based misuse-detection and anomaly detection using heuristic methods, respectively. This paper proposes hybrid IDS with a soft computing method that serves two folds: First, the hybrid IDS method is obtained by combining packet header anomaly detection (PHAD) and application header anomaly detector (AHAD) which are used to analyze anomaly-based IDSs. Second, the work presents an IDS alarm classifier based on the Fuzzy C-Means clustering algorithm to cluster traffic flow patterns. The proposed combined techniques of Gravity search algorithm and Gravity wolf optimization with deep neural network (GSGW-DNN) automatically classify normal or malicious attacks. The experimental result for the overall proposed system performance are evaluated with high accuracy, precision, recall, and F-measure and compared with existing ANN and SVM algorithms. The evaluation results prove that the proposed approach is suitable for effective attack detection and prevention.