AdaptiveShield: Dynamic Defense Against Decentralized Federated Learning Poisoning Attacks

Summary

AdaptiveShield is a hybrid defense for decentralized federated learning that targets poisoning attacks, in which malicious participants inject false data to corrupt the global model. The approach uses dynamic detection that weighs multiple risk factors to compute a maliciousness index and adjust detection thresholds, and it adapts hyperparameters to limit damage from missed attackers. The authors report over 0.1 improvement in training accuracy over state-of-the-art approaches with negligible time overhead.

Related items

• InfoSecuring next-generation edge intelligence: A survey from federated learning to autonomous agents in cyber-physical systemsSimilar attack · OpenAlex (peer-reviewed AI security)
• InfoPop Goes the Verifier: Phoneme-Targeted Data Poisoning of Voice Liveness DetectionSimilar attack · Arxiv (cs.RO + cs.CV security)
• InfoCharacterizing Recoverable Expert-Utilization Signatures in Fine-Tuned Mixture-of-Experts LLMsSimilar attack · Arxiv (cs.CR + cs.CL + cs.LG)
• InfoDiscrete Unintentional Adversarial Perturbation Detection for DQN Traffic Signal ControlSimilar attack · OpenAlex (peer-reviewed AI security)
• InfoTrustworthy Distributed Machine Learning in Cloud-Edge-IoT Systems: A Heterogeneity-Aware Survey of Security, Privacy, and RobustnessSimilar attack · Arxiv (cs.CR + cs.CL + cs.LG)