Hybrid Optimized 3D Localization for WSN-Assisted IoT Networks in Smart Agriculture

Accurate localization of sensor nodes within precision agriculture applications is a critical component in Wireless Sensor Network (WSN)-assisted Internet of Things (IoT) networks. The presence of environmental noise, terrain irregularities, and data anomalies degrades the performance of the existing 3D localization techniques. This article presents a novel hybrid 3D localization scheme that integrates Particle Swarm Optimization (PSO), Random Forest (RF)-based anomaly detection, and trilateration refinement to enhance localization accuracy, energy efficiency, and scalability in smart agriculture (SA) environments. The proposed scheme proceeds in three phases, i.e., initial node position estimation using RSSI-based path-loss modelling, machine learning (ML)-based anomaly detection and filtering of Received Signal Strength Indicator (RSSI) data, and PSO-based global optimization followed by trilateration for fine-tuning. Based on the simulation experiments for several scenarios, the proposed hybrid approach renders a robust and scalable solution for accurate node localization in WSN-assisted IoT (WIoT) networks for smart agriculture. It attains low localization errors at 1.2 meters, with energy consumption abridged to 8–10 J and computation time under 0.5 seconds, outdoing the state-of-the-art.