Synthetic dataset for the study of wireless propagation modeling through a Bayesian network

Synthetic dataset for the study of wireless propagation modeling through a Bayesian network, which is composed of all the variables involved in the urban versions of the following path loss models: Okumura-Hata, Lee (Philadelphia, Newark, and Tokyo cases), and COST 231-Walfish Ikegami. A subset of four input variables was common for all models: Hbs (base station antenna height, in meters), Hms (mobile station antenna height, in meters), d (distance from base station, in kilometers), and f (carrier frequency, in megahertz). Two additional input variables were required by the COST 231-Walfish Ikegami model, namely b (distance between buildings, in meters) and nf (average number of floors per building). Together with the output variables corresponding to the five path loss calculations, the dataset consisted of a total number of 11 variables organized by columns. Then, we generated each row by randomly sampling each input variable according to a uniform distribution, and introducing the resulting path losses. In total, we randomly generated 10, 000 observations.