IoT Security Model with Machine Learning and Edge Computing for Smart Poultry Farm

This study discusses the application of the Internet of Things (IoT) in the poultry farming sector, which is increasingly advancing through the integration of machine learning and edge computing to improve production efficiency and animal welfare. However, the cybersecurity aspect of this system remains a major challenge. The study aims to design an IoT security model based on machine learning and edge computing that can detect and prevent anomalies or potential cyberattacks in real-time within smart poultry farming systems. The developed model utilizes environmental and operational data from IoT sensors, which are processed locally using edge devices and analyzed with the Random Forest algorithm for early detection of suspicious activities. In addition to Random Forest, the performance evaluation also involves benchmarking algorithms such as Support Vector Machine (SVM), Multi-Layer Perceptron (MLP), and Extreme Gradient Boosting (XGBoost). Initial results show that the combination of Random Forest architecture and edge computing provides the highest anomaly detection accuracy and the lowest processing latency compared to other models, emphasizing the importance of a proactive, artificial intelligence-based security approach in modern agricultural IoT environments.