arXiv:2508.15764cs.LGcs.MA2025-08中稿 · publication at ECA…被引 1

提出一种无需全局信息的分布式检测方法,实时识别连续动作多智能体中的对抗攻击。

Distributed Detection of Adversarial Attacks in Multi-Agent Reinforcement Learning with Continuous Action Space

  • 基于本地观测构建智能体正常行为的高斯分布模型,实现去中心化检测。
  • 在多个PettingZoo环境中对主流攻击方法检测准确率超0.95 AUC-ROC。
  • 适用于需要实时安全防护的协作式多智能体系统,如自动驾驶车队。

本文针对具有连续动作空间的协作式多智能体强化学习中的对抗攻击检测问题,提出一种去中心化检测器。该检测器仅依赖智能体的本地观测,通过深度神经网络将可观测智能体的正常行为建模为参数化的多元高斯分布。基于预测的密度函数,定义正常性得分,并给出其均值与方差的刻画。据此采用双侧CUSUM过程检测正常性得分偏离均值的情况,实现实时异常行为识别。我们在多种多智能体PettingZoo基准测试中评估该方案,针对不同前沿攻击方法,结果表明该方法能有效检测具有显著影响的对抗攻击。尤其在所有测试环境中,对最具破坏性的攻击检测性能优于离散对应方法,AUC-ROC得分超过0.95。

原文摘要 · Abstract (English)

We address the problem of detecting adversarial attacks against cooperative multi-agent reinforcement learning with continuous action space. We propose a decentralized detector that relies solely on the local observations of the agents and makes use of a statistical characterization of the normal behavior of observable agents. The proposed detector utilizes deep neural networks to approximate the normal behavior of agents as parametric multivariate Gaussian distributions. Based on the predicted density functions, we define a normality score and provide a characterization of its mean and variance. This characterization allows us to employ a two-sided CUSUM procedure for detecting deviations of the normality score from its mean, serving as a detector of anomalous behavior in real-time. We evaluate our scheme on various multi-agent PettingZoo benchmarks against different state-of-the-art attack methods, and our results demonstrate the effectiveness of our method in detecting impactful adversarial attacks. Particularly, it outperforms the discrete counterpart by achieving AUC-ROC scores of over 0.95 against the most impactful attacks in all evaluated environments.

多智能体对抗攻击检测强化学习

Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品;中文卡片由大模型生成,请以原文为准。