The method to generate a combined dataset of electromagnetic environment sensing data and topographic data for the prediction of radio wave propagation loss

Abstract

Since the topography of mountains is complex, it is difficult to obtain topographic data related to longitude, latitude and elevation and the spectrum data related to field strength, thus causing the difficulty in establishing the dataset used to build the model of radio wave propagation loss. This paper put forward a method to generate a combined dataset of electromagnetic environment sensing data and topographic data to predict the loss of radio wave propagation. By using a software named LocaSpace Viewer, we got the data on longitude, latitude and elevation between the radiant points and measuring points. Also, the software of Winprop was used to generate the simulated sensing data of electromagnetic environment according to their corresponding geographic positions. With all these efforts, we designed an algorithm to better select the data and obtain the combined dataset of the sensing data in specific electromagnetic environment and the topographic data under the UTM coordinate. We input the dataset and trained them to get the radio wave propagation loss model in random forests. After the testing, we demonstrated that the random forests model built under the help of the dataset had a good performance in calculating radio wave propagation loss.