Feature Extraction Using Hough Transform in Radio Propagation Estimation

This paper proposes a feature extraction method using the Hough transform as a method for utilizing map images in radio propagation estimation. The regression results of propagation loss parameters are evaluated in comparison with those by conventional methods using CNNs. Furthermore, we discuss how to extract feature parameters effectively.

I. Introduction

Accurate and rapid understanding of a radio propagation environment is a mandatory technique for high-speed wireless communication. The path of radio waves often becomes multipath, where transmitted waves are repeatedly reflected and diffracted, and reach mobile stations passing through different and complex paths. Many methods have already been proposed for predicting radio propagation environments that can be classified into empirical and deterministic methods [1].