Mis-spoke or mis-lead: Achieving Robustness in Multi-Agent Communicative Reinforcement Learning
arXiv:2108.03803
Abstract
Recent studies in multi-agent communicative reinforcement learning (MACRL) have demonstrated that multi-agent coordination can be greatly improved by allowing communication between agents. Meanwhile, adversarial machine learning (ML) has shown that ML models are vulnerable to attacks. Despite the increasing concern about the robustness of ML algorithms, how to achieve robust communication in multi-agent reinforcement learning has been largely neglected. In this paper, we systematically explore the problem of adversarial communication in MACRL. Our main contributions are threefold. First, we propose an effective method to perform attacks in MACRL, by learning a model to generate optimal malicious messages. Second, we develop a defence method based on message reconstruction, to maintain multi-agent coordination under message attacks. Third, we formulate the adversarial communication problem as a two-player zero-sum game and propose a game-theoretical method R-MACRL to improve the worst-case defending performance. Empirical results demonstrate that many state-of-the-art MACRL methods are vulnerable to message attacks, and our method can significantly improve their robustness.
arXiv commentPublished as a conference paper in AAMAS 2022
References in corpus (11)
- PyTorch: An Imperative Style, High-Performance Deep Learning Library
- Learning to Communicate with Deep Multi-Agent Reinforcement Learning
- Learning Attentional Communication for Multi-Agent Cooperation
- Social Influence as Intrinsic Motivation for Multi-Agent Deep Reinforcement Learning
- Building Generalizable Agents with a Realistic and Rich 3D Environment
- QPLEX: Duplex Dueling Multi-Agent Q-Learning
- Guided Deep Reinforcement Learning for Swarm Systems
- Learning Nearly Decomposable Value Functions Via Communication Minimization
- A Generalized Training Approach for Multiagent Learning
- The Emergence of Adversarial Communication in Multi-Agent Reinforcement Learning
- Gaussian Process Based Message Filtering for Robust Multi-Agent Cooperation in the Presence of Adversarial Communication