(英) |
Various propagation loss models have been proposed for the design of base stations in mobile communication systems. The most well-known methods for creating propagation loss models are multiple regression analysis based on multiple measurement results and reflection/diffraction theory-based methods. In general, the estimating equations obtained by these methods have a problem of limited applicability. In this paper, we propose a method for estimating the propagation model using machine learning. In order to improve the accuracy of the model for propagation loss estimation using machine learning, we need to consider a wide range of issues, such as selecting an appropriate model for the problem, adjusting hyperparameters, and so on. One of the most important issues is how to appropriately calculate the input parameters for the estimation target. In this paper, we focus on the method of obtaining appropriate input parameters with the aim of improving the accuracy of propagation loss estimation. |