Analytical Expression for Recommendation ITU-R P.l546-6 Propagation Curves of Land Paths Up to 20 km Using Symbolic Regression

Path loss is a fundamental parameter in the planning of radio systems, such as wireless loT and 5G networks. In the scope of propagation models for link budget analysis of such systems, the ITU-R P.1546-6 Recommendation model stands out as a suitable approach, in which path loss is obtained based on measured electric field strength data. It can be utilized for the design of terrestrial services such as broadcasting, maritime mobile, land mobile, and point-to-multipoint systems operating in the frequency range 30 MHz to 4 GHz. The model is mainly based on the use of propagation curves derived from those empirical data, combined with an intricate sequence of interpolations. A possible way to enhance the model efficiency is the introduction of concise and precise analytical expressions to replace part of those interpolations. This paper utilizes a data-driven machine learning tool known as symbolic regression for extracting governing analytic equations from part of the propagation curves in the Recommendation. By using such method, novel concise formulas are proposed to predict the electric field strength variation with distance and base antenna height, instead of interpolating data, for land paths up to 20 km. The proposed expressions present maximum error less than 2 dB with respect to the Recommendation data.