A Collaborative Approach to Detect DDoS Attacks in OpenStack-based Cloud using Entropy and Machine Learning

Cloud computing involves optimizing the efficiency of data storage and processing through the use of networking, various software, and internet services. In this context, network security is of utmost importance to prevent cybercriminals from disrupting resources and impeding effective utilization. Among the primary security concerns in cloud computing are Distributed Denial of Service (DDoS) attacks, which can have severe consequences for business continuity. To address this, it is crucial to develop a robust DDoS detection system that can identify suspicious activities within the cloud environment. In this study, we propose a collaborative approach that combines entropy and machine learning techniques to detect DDoS attacks specifically in an OpenStack-based cloud environment. Initially, we employ an entropy-based technique to identify vulnerable hosts, followed by a machine learning-based technique to identify attacker hosts. To train and evaluate our detection system, we utilize real-time cloud data obtained from an OpenStack-based cloud testbed. Results reveal that using logistic regression improves the accuracy of the classifier. Furthermore, the collaborative approach significantly reduces the detection time.