Improved Multiview Graph Convolutional Network for Threat Detection in Internet of Things-based Smart Farming Systems
S. P. Prakash, Haydeer MohamadAbbas, A Meghana, A. G., K. Ghamya · 2025
Over past few years, the agriculture industry has experienced a significant transformation in farming by adapting advance technologies like Internet of Things (IoT), which leads to a modern era of smart farming systems. These systems use smart technologies such as IoT devices and sensors for optimize farming. However, these systems possess various cyber threats such as data breaches, device tampering and unauthorized access which complicates the system functionality. Therefore, this research proposes Improved Multiview Graph Convolutional Network (IMGCN) for threat detection. Initially, data is collected from Neuro Fuzzy Bot (NF-Bot-IoT) and preprocessed by using Synthetic Minority Over-sampling Technique (SMOTE) which handles class imbalance issues. Further, feature selection is performed using Recursive Feature Elimination (RFE) which reduces the impact of noisy or irrelevant features. Finally, threat detection in smart farming systems is done by using proposed IMGCN which detects threats and protects data. The proposed IMGCN achieved better results in terms of accuracy (93.25%), precision (92.23%) as well as F1-Score (91.78%) when compared with existing Multi-Layer Perceptron (MLP).