Anomaly Detection in Smart Agriculture Systems on Network Edge Using Deep Learning Technique

Abstract

With the widespread adoption of Internet of Things (IoT) technologies across various domains, including smart agriculture, urban environments, and homes, the threat of zero-day attacks has surged. This research delves into the application of deep learning techniques to detect anomalies in smart agricultural systems at the network edge, with a specific focus on safeguarding them against Distributed Denial of Service (DDoS) attacks. In this study, we propose an anomaly detection model based on CNN-LSTM to analyze sensor data collected from IoT devices. We rigorously train and test our model using two distinct datasets of sensor readings, simulating potential DDoS attack scenarios. The model's performance is assessed using key metrics such as detection accuracy, recall, and F1-score. Our results demonstrate the effectiveness of our approach, achieving an impressive anomaly detection accuracy of 99.7%. This research contributes significantly to the development of robust and efficient attack and anomaly detection techniques for smart agriculture systems at the network edge, ultimately enhancing the reliability and sustainability of agricultural practices.