Leveraging OpenStack-based Private Cloud for Intrusion Detection Using Deep Learning Techniques

OpenStack is a cloud platform that offers infrastructure services, providing access to computing, storage, and network resources. However, due to its open-source nature, the cloud environment can be vulnerable to security threats. OpenStack security measures are generally applied to specific services, rather than being integrated comprehensively across the entire cloud environment. Despite offering flexibility and scalability, the widespread use of OpenStack calls for robust security solutions. Current intrusion detection systems often struggle to adapt to constantly evolving traffic patterns and attack vectors in the OpenStack environment. This study introduces a two-stage deep learning framework to create a robust Intrusion Detection System (IDS) for OpenStack-based private clouds. The proposed framework efficiently analyzes network traffic data in these clouds by implementing dimensionality reduction and anomaly detection techniques. It utilizes a network-based IDS, trained on two publicly available datasets, which enables a broader detection range of known attack patterns. The results indicate that the OpenStack IDS significantly enhances OpenStack's security, addressing the limitations of generic IDS solutions.