Abstract
Positioning is a fundamental process in wireless sensor networks (WSNs) and Internet of Things (IoT) systems to accurately track the location of nodes. This study presents a cost-effective and scalable single-anchor node positioning system integrating ZigBee technology, directional antennas, and machine learning (ML). The system utilizes a custom-designed directional antenna array and the XGBoost algorithm to process Received Signal Strength Indicator (RSSI) data, achieving precise localization with minimal hardware requirements. Experimental results confirm the superiority of the antenna array, demonstrating a maximum gain of 10 dBi and enhanced signal reception compared to monopole and single-patch antennas. The XGBoost model achieved high accuracy, with R² values of 97.76% on training and 90.95% on test data.