Low-Cost Path Loss Estimation Using Correlation Graph CNN with Novel Feature Parameters

In this paper, we propose a method for predicting radio wave propagation using a correlation graph convolutional neural network (C-Graph CNN). We examine what kind of parameters are suitable to be used as system parameters in C-Graph CNN. Performance of the proposed method is evaluated by the path loss estimation accuracy and the computational cost through simulation.

I. Introduction

The radio wave propagation environment needs to be accurately predicted when designing and selecting base stations for mobile communication systems to achieve high-speed communication. Empirical and deterministic models are two typical methods used to predict radio wave propagation [1]. The empirical model is obtained by statistically processing and formulating the measurement data, whose examples include the Okumura-Hata model [2] and the Walfisch-Ikegami model [3], [4]. On the other hand, a deterministic model is used to estimate radio propagation characteristics on the basis of electromagnetic field theory, such as the finite-difference time-domain (FDTD) method [5] and ray tracing method [6], [7].