Оптимізація розподілу ресурсів у хмарних серверних системах на основі методів нечіткої логіки

Abstract

This research addresses inefficient resource utilization and SLA noncompliance caused by volatile traffic in cloud environments. We propose a fuzzy logic-based control mechanism that evaluates real-time CPU and memory metrics to optimize scaling decisions. Unlike rigid threshold-based strategies, the fuzzy inference system provides fluid transitions, minimizing instability in virtual machine management. Simulations demonstrate that this approach boosts energy efficiency by 15% while sustaining high availability