Foundation Model-aided Multi-agent Reinforcement Learning for Random Access Network Optimization

Random access (RA) is one of the most foundational medium access control (MAC) layer scheduling schemes for handling unpredictable data traffic from multiple terminals and serves as the basis for modern carrier-sense multiple access (CSMA) protocols. While multi-agent reinforcement learning (MARL) has been explored to optimize RA-based networks, its reliance on experience-driven, distributed policy learning incurs significant training overhead for each optimization task, limiting their feasibility in real-world applications. In this work, we propose to leverage a foundation model (FM) to improve MARL efficiency across diverse RA network optimization tasks. Specifically, we design an FM-aided actor-critic algorithm within a consensus-based decentralized MARL architecture and provide its convergence analysis. Numerical evaluations show that our proposed method enhances MARL efficiency for RA network optimization.