Prediction of Communication Signal Strength with UAVs Using Artificial Neural Networks

Jaqueline dos Santos Silva, Evelio Fernandez, Alessandro Zimmer

DOI: 10.14209/sbrt.2024.1571026178

Evento: XLII Simpósio Brasileiro de Telecomunicações e Processamento de Sinais (SBrT2024)

Keywords: uav artificial neural networks path prediction signal strength

Abstract

Recognizing the growing importance of unmanned aerial vehicles in urban traffic surveillance, this research aims to predict Wi-Fi signal strength during drone flights. A multilayer perceptron was employed and achieved an RMSE of 0.47. In comparison, signal strength predictions using the Longley-Rice model through Radio Mobile presented higher error metrics, with RMSE values ranging from 8.23 in rural areas to 12.84 in urban scenarios, highlighting that artificial neural networks are a promising methodology for predicting signal strength.

Download