DSN: A DDPG-Based Scheduling Framework for Optimal Task Allocation in Cloud Data Centers

In modern cloud data center environments, precise and efficient scheduling of data processing jobs emerges as paramount. While many production environments conventionally adopt heuristic strategies for scheduling batch processing jobs, these methods may fall short of achieving optimal scheduling in the face of complex environments or specialized tasks, thereby significantly extending the completion time of jobs. To address this issue, we introduce the DSN (Deep-Schedule-Net) framework—a cluster batch processing job scheduling method based on the DDPG algorithm and tailored to address task scheduling optimization problems under multi-resource constraints. Compared to other related algorithms, experimental results substantively affirm that the proposed algorithm markedly optimizes in curtailing batch task completion time.