Security-Aware Cloud Environment Modelling and Resource Profiling for Intelligent Cloud Resource Management

Abstract

The quick extension of cloud computing has yielded in emergingly diverse methodologies, adaptable workloads, and complicated interaction ecosystems, generating difficulties for effective and secure resource management. Traditional cloud resource management methodologies primitively consider computational resource accessibility and usage, while security-relevant characteristics are generally treated independently. The proposed Integrated Security-Aware Cloud Resource Profiling Framework that combines computational, workload, network, and security characteristics into a unified cloud resource profile. The proposed approach models diverse cloud frameworks utilizing CloudSim Plus and includes network behavior via NS-3. Diverse workloads are characterized as per the memory, processing, bandwidth, execution, priority, latency, and security essentials. VMs and physical hosts are consequently profiled based on resource capacity, workload allotment, utilization, security characteristics and operational conditions is performed at workload, VM, and host levels utilizing access nature, resource threats, interaction trends, policy compliance and security events. The developed frame accomplished 98.3%, 97.2%, 96.1% and 96.7% of accuracy, precision, recall, and F1-score respectively. For security profiling with 18.5%, 97.8% and 12.4ms of resource usage, profiling accuracy, and mean network latency respectively. The resulting hybridized security-aware cloud model facilitates a systematic base for consequent smart resource allotment and orchestration.