Adaptive On-the-Fly Scan Method for Fast and Efficient Planar Near-Field Acquisition

Recent advancements in adaptive planar near-field scanning have enhanced the use of sparse measurement data for electromagnetic field (EMF) modeling and prediction. In this work, we propose an adaptive planar scanning solution that utilizes iterative row-by-row scans, where each row is scanned on-the-fly (OTF). Both the total number of rows and their separation are dynamically adjusted and optimized through correlation analysis of the values obtained from previous measurements. For initialization, a few evenly distributed row scans, typically 4-6 OTF scans, are performed, followed by an iterative process of finding new rows to scan using Gaussian Process Regression (GPR) based Bayesian optimization. For validation, both numerical and experimental examples, including wire and patch antenna arrays, are provided. Results show good performance of the proposed method, enabling rapid near-field acquisition within minutes, making it suitable for large-scale and high-resolution EMF scanning applications.