Research on Edge Computing and Cloud Collaborative Resource Scheduling Optimization Based on Deep Reinforcement Learning
Year: 2025
Type: conference-paper
Abstract: Resource scheduling optimization in edge-cloud collaborative computing is a critical challenge due to dynamic workloads, latency constraints, and limited edge resources. This study proposes a deep reinforcement learning (DRL)-based scheduling approach to enhance task processing efficiency, minimize overall processing time, optimize resource utilization, and control task migrations. Experimental results demonstrate that the proposed DRL model outperforms traditional scheduling algorithms, reducing total processing time by up to 18% and improving resource utilization by 12% under high task loads. Furthermore, DRL effectively balances task allocation, leading to 30% fewer task migrations compared to priority-based and load-balancing scheduling methods. Despite these improvements, challenges remain in learning efficiency, training overhead, and convergence stability. Future research should focus on enhancing algorithm fault tolerance and accelerating training convergence to handle large-scale and uncertain scheduling environments, further advancing intelligent resource management in edge-cloud computing systems.
Authors: Yuqing Wang, Xiao Yang
Language: English
Topic: IoT and Edge/Fog Computing
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W4411143685