Published Paper Details:

Distributed Denial of Service (DDoS) attacks pose a significant threat to cloud computing infrastructures, with the potential for substantial financial damage. Current detection approaches often fall short in addressing the sophisticated attack strategies that exploit the unique characteristics of cloud systems, such as elasticity and multi-tenancy. This research proposes a novel framework that leverages trust-based mechanisms to improve the detection of suspicious nodes within cloud-based virtual machines. Through combining data-driven and opinion-based trust sources via Bayesian reasoning, the hypervisor can establish reliable relationships with guest VMs and make informed decisions about their trustworthiness. Additionally, a game-theoretic approach is introduced to optimize the allocation of detection load among VMs, maximizing the identification of DDoS attacks while minimizing the utilization of critical cloud resources. Comprehensive evaluations demonstrate the effectiveness of the proposed solution, with enhanced attack detection rates, reduced false positives and negatives, and efficient resource utilization compared to existing methodologies.