Federated Autonomous Cyber Agents for Privacy-Preserving Intrusion Detection in Distributed Smart Grids
Abstract
Abstract Distributed smart grids increasingly rely on interconnected cyber-physical infrastructures, edge intelligence, microgrids, distributed energy resources, and real-time demand-response mechanisms. Although this integration improves flexibility and operational efficiency, it also exposes power systems to false data injection, denial-of-service, replay, command manipulation, poisoned model updates, and privacy-inference attacks. Conventional centralized intrusion-detection systems require sensitive grid data to be transferred to a central server, while standard federated learning approaches remain vulnerable to non-IID data, unreliable agents, Byzantine updates, and weak operational response capability. To address these limitations, this paper proposes a federated autonomous cyber-agent framework for privacy-preserving intrusion detection in distributed smart grids. The proposed model integrates local temporal intrusion detection, hybrid classification and anomaly scoring, differentially private update protection, secure aggregation, trust-weighted Byzantine-resilient filtering, and grid-safe autonomous response. Each local cyber agent detects cyber-physical threats using private domain data, shares only protected model updates, and selects mitigation actions under stability, uncertainty, and human-override constraints. Simulation-based evaluation using smart-grid, IIoT, and network-intrusion benchmark settings indicates that the proposed framework improves detection…
Read the paper
Related studies
- Active Learning during Federated Learning for Object Detection
- Fairness-aware federated learning
- Decentralized Federated Learning
- Building a Personally Identifiable Information Recognizer in a Privacy Preserved Manner Using Automated Annotation and Federated Learning
- Preservation of Patient Level Privacy: Federated Classification and Calibration Models
- Efficient explainable federated learning
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.