Machine learning - powered intrusion detection system for agriculture 4.0: Securing the next generation of farming
Abstract
Agriculture is a fundamental component of India's economy, involving more than half of the workforce and significantly contributing to revenue generation via exports. Integrating IoT, drones, and artificial intelligence within Agriculture 4.0 transforms conventional farming methods, enhancing productivity and sustainability. However, the integration of IoT devices in open fields poses considerable cybersecurity issues, including susceptibility to Distributed Denial of Service (DDoS) attacks and the potential for data manipulation. These assaults undermine the integrity, reliability, and safety of agricultural systems, underscoring the pressing need for stringent security measures. Recent research underscores the efficacy of intrusion detection systems (IDS) in detecting and mitigating these attacks. This study tackles existing gaps by introducing a meta-model that integrates XGBoost (XG), Random Forest (RF), Decision Trees (DT), and a meta-learner utilizing XGBoost. The model demonstrated impressive results, achieving an accuracy of 96.37%, precision of 96.39%, recall of 96.37%, and an F1 score of 96.25%. These findings illustrate the model's effectiveness in ensuring Agriculture 4.0, which contributes to resilient and sustainable innovative farming ecosystems.
Authors

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.