FULL PDF
Abstract
The rapid adoption of cloud computing has significantly amplified energy consumption, necessitating sustainable and energy-efficient solutions. This paper presents a reinforcement learning-based adaptive workload balancing mechanism tailored for energy-aware task offloading in green cloud computing environments. We analyze how dynamic workload shifts, when guided by intelligent policies, reduce power usage while maintaining service quality. The paper also surveys key contributions made before 2022 and illustrates optimization models, datasets, and visual comparisons of task offloading strategies. The findings highlight that reinforcement learning (RL) algorithms can effectively learn energy-optimal offloading policies under real-time constraints.
View more >>