AI ‐Driven Resource Optimization in Distributed Computing for Hybrid Cloud Environments: An Energy‐Aware Approach

ABSTRACT

Optimization of resources in hybrid cloud‐edge computing systems poses great computational/economic issues. The conventional rigid allocation schemes cannot keep up with the dynamic workload characteristics and lead to energy wastage and excess cost. This study presents a framework of AI‐based optimization as a collaboration between deep reinforcement learning (DRL) and quantum‐inspired mechanisms of allotment of assets dynamically in hybrid clouds. We test our design with 487 workload instances of a representative workload and on heterogeneous computing infrastructure consisting of edge nodes, fog computing layers, and cloud data centers. Our quantum‐inspired genetic algorithm attains a 42% reduction in energy consumption over the baseline techniques with a service level agreement (SLA) of 98.7%. The researchers apply a deep Q‐network (DDQN) agent to improve costs by 38% with smart scaling choices. Technical validation will show that our hybrid DRL‐quantum solution will save USD 47250 monthly (39% savings) and will lead to the improvement of resource utilization (64.2–88.1). We offer detailed algorithmic descriptions, verified experimental findings, and deployment plans of cloud operators of heterogeneous computing resources over distributed infrastructure.