Advancing Multi-Agent Systems Integrating Federated Learning with Deep Reinforcement Learning: A Survey

Recent advancements in the field of Deep Reinforcement Learning (DRL), such as the development of algorithms like Deep Q-Network, Deep Deterministic Policy Gradient (DDPG), and Proximal Policy Optimization (PPO), have been significant. However, these algorithms encounter data privacy issues in multi-agent environments. This paper investigates the latest research trends in applying federated learning to address data privacy issues in multi-agent systems within deep reinforcement learning. It delves into various approaches for implementing federated learning in these systems, examines the challenges faced, and explores potential solutions to enhance privacy while maintaining or improving the performance of DRL algorithms in multi-agent setting.