An Intrusion Detection System in Cloud Environments using Combined Machine Learning Algorithms

One of the major security issues in cloud computing is protection against network intrusions that affect confidentiality, availability and integrity of cloud resources and offered services. Since, cloud computing is a shared environment it is therefore subject to various attacks. Due to increasing incidents of cyber attacks, building effective intrusion detection systems are essential for protecting the cloud environments security, and till now it remains an elusive goal and a big challenge. In this paper, an Intrusion Detection System (IDS), regarding the host-based model for cloud computing is proposed according to the Bagging algorithm for virtual machines. In the pre-processing stage, by using a feature selection algorithm, we have chosen 11 of 11 features from the NSL-KDD dataset. We have used three well-known classifiers, namely the decision tree, Bayesian and K- Nearest Neighbor (KNN). Simulation results compared with these three classifiers separately and another method using the Random Forest algorithm shows improvements in terms of precision, recall, TPR and FPR.

Keywords:

Intrusion Detection System , Bagging , Bayesian classifier , KNN , C4.5 algorithm