Cloud-Fog-Edge Collaborative Computing for Sequential MIoT Workflow: A Two-Tier DDPG-Based Scheduling Framework

2025-10-23

AI Summary

This work addresses the NP-hard, end-to-end latency-sensitive scheduling problem for sequential medical Internet-of-Things (MIoT) workflows in heterogeneous cloud–fog–edge environments, where the objective is to minimize makespan. We propose a cloud–fog–edge collaborative hierarchical reinforcement learning (HRL) scheduling framework. Our approach features a novel two-level Deep Deterministic Policy Gradient (DDPG) architecture that decouples high-level “tier selection” from low-level “node assignment,” enabling joint global resource optimization and local dynamic adaptation. A hierarchical controller mechanism is further introduced to support cross-domain, long-horizon policy learning. Experimental results on complex MIoT workflows demonstrate that our framework significantly reduces makespan—achieving up to a 23.6% improvement over state-of-the-art baselines—while exhibiting strong scalability and real-time adaptability to dynamic environment changes.

Technology Category

• Planning, Routing, and Scheduling: Planning/Scheduling and Learning
• Machine Learning: Hardware-aware ML
• Reasoning under Uncertainty: Sequential Decision Making

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

The Medical Internet of Things (MIoT) demands stringent end-to-end latency guarantees for sequential healthcare workflows deployed over heterogeneous cloud-fog-edge infrastructures. Scheduling these sequential workflows to minimize makespan is an NP-hard problem. To tackle this challenge, we propose a Two-tier DDPG-based scheduling framework that decomposes the scheduling decision into a hierarchical process: a global controller performs layer selection (edge, fog, or cloud), while specialized local controllers handle node assignment within the chosen layer. The primary optimization objective is the minimization of the workflow makespan. Experiments results validate our approach, demonstrating increasingly superior performance over baselines as workflow complexity rises. This trend highlights the frameworks ability to learn effective long-term strategies, which is critical for complex, large-scale MIoT scheduling scenarios.

Problem

• Minimizing makespan for sequential MIoT workflows
• Addressing NP-hard scheduling in cloud-fog-edge systems
• Optimizing hierarchical resource allocation for healthcare workflows

Innovation

• Two-tier DDPG framework for hierarchical scheduling
• Global controller performs layer selection
• Local controllers handle node assignment