Intelligent Service Chain Orchestration and Resource Allocation in End-Edge Collaborative IIoT Using Multi-Agent...

Area of Science:

• Computer Science
• Network Engineering
• Artificial Intelligence

Background:

• Industrial Internet of Things (IIoT) networks generate massive data streams requiring low-latency processing.
• Edge network resource management faces challenges due to heterogeneous data and stringent performance demands.
• Existing solutions struggle with the complexity of joint service function chain orchestration and resource allocation.

Purpose of the Study:

• To address the joint optimization of Service Function Chain (SFC) orchestration and resource allocation in edge gateway-assisted IIoT networks.
• To minimize end-to-end latency and energy consumption while ensuring Quality of Service (QoS) constraints.
• To develop an efficient algorithm for a complex, NP-hard optimization problem in distributed environments.

Main Methods:

• Formulated the problem as a mixed-integer nonlinear programming (MINLP) model.
• Proposed the SFC Orchestration and Resource Allocation-based Multi-Agent Proximal Policy Optimization (SORA-MAPPO) algorithm.
• Implemented a centralized training with decentralized execution (CTDE) paradigm incorporating intelligent agent cooperation.

Main Results:

• The SORA-MAPPO algorithm effectively handles the NP-hard problem of SFC orchestration and resource allocation.
• The proposed scheme demonstrates significant improvements in minimizing latency and energy consumption.
• Simulations confirmed the algorithm's effectiveness in complex and dynamic IIoT scenarios.

Conclusions:

• The SORA-MAPPO algorithm provides an efficient solution for resource management in edge-assisted IIoT networks.
• The CTDE approach with agent cooperation is suitable for distributed optimization tasks with partial observability.
• This work contributes to enhancing the performance and efficiency of IIoT systems.