A study on path loss prediction based on Tikhonov regularization using gradient adaptive step‐size in mountain areas

Abstract

AbstractThis paper presents a method of applying a regularization concept for more accurate path loss prediction in mountainous areas. Tikhonov regularization with gradient adaptive step‐size is used to adjust the path loss model. Measurements were conducted in two mountainous areas to compare the path loss prediction performances. Prediction accuracies were improved by up to 3.42 and 0.95 dB in the average error and the SD of errors by the proposed method when compared to the nontuned log‐distance path loss model and Edwards‐Durkin model with diffraction models.