A CLUSTER-BASED KEY GENERATION MODEL FOR INITIATING AN EFFICIENT CLOUD AUDITING SYSTEM

A cloud provider's transparency and accessibility to its tenants is ensured by cloud security audits. Due to the intricate operational demands associated with multi-tenancy and self-service features of cloud computing, along with the substantial scale of cloud environments, security audits in the cloud can be prohibitively costly and lack scalability. Consequently, one possible approach to achieving a satisfactory response time is to adopt a proactive auditing technique that initiates the auditing process prior to critical events. Nonetheless, a significant drawback of these methods is their dependence on manual labor to derive the interdependencies between occurrences, thereby limiting their applicability and flexibility. Our research introduces clustering-based key generation model where an entirely automated system that builds dependency models from runtime events using learning-based techniques, enabling proactive security auditing of cloud activities. Our findings highlight notable benefits, such as a roughly 60% improvement in speed compared to current proactive methods, and a feasible response time of 6ms for auditing a cloud. We do thorough testing in real and virtual cloud scenarios, and we integrate clustering with the popular cloud platform.