Emergent Intrusion Detection System for Fog Enabled Smart Agriculture Using Federated Learning and Blockchain Technology: A Review

Abstract

The rapid evolution of smart agriculture has revolutionized traditional farming practices by integrating Internet of Things (IoT), artificial intelligence (AI), and fog computing to enhance productivity, efficiency, and sustainability. However, this increasing interconnectivity also exposes smart agricultural systems to cybersecurity vulnerabilities, necessitating robust and adaptive Intrusion Detection Systems (IDS). This paper presents a comprehensive review of the latest advancements in intrusion detection for fog-enabled smart agriculture, focusing on the synergistic integration of federated learning (FL) and blockchain (BC) technologies. FL enables collaborative privacy-preserving anomaly detection across distributed agricultural IoT nodes, mitigating data exposure risks. Meanwhile, blockchain strengthens security by providing decentralized trust management, immutable logging, and secure model aggregation in FL-based IDS. We analyze existing state-of-the-art approaches, highlight their advantages and limitations, and discuss emerging challenges, such as adversarial attacks, computational overhead, data heterogeneity, and communication constraints in FL-based IDS frameworks. Furthermore, we examine how blockchain enhances the resilience of federated learning against security threats while maintaining system integrity in real-world smart farming applications. This review also proposes a novel system architecture that optimally integrates fog computing, federated learning, and blockchain to enhance intrusion detection accuracy, energy efficiency, and system resilience in smart agriculture. The insights provided in this review aim to guide researchers and practitioners in developing next-generation, secure, and adaptive intrusion detection frameworks for future cyber-resilient smart agriculture.