Técnicas inteligentes para identificação de linha de visada e alocação de fator de espalhamento em redes LoRaWAN

Abstract

This work investigates intelligent strategies to optimize performance and to reduce energy consumption in LoRaWAN networks by using Reinforcement Learning for the dynamic allocation of spreading factors, based on the identification of line-of-sight conditions between network nodes. For this, this work incorporates a real-world measurement campaign conducted at the Federal University of Juiz de Fora (UFJF) alongside simulations using the ns-3 simulator to validate the proposed strategies. The results indicate that the integration of channel modeling techniques and the adoption of reinforcement learning enable a more efficient allocation of transmission parameters, yielding significant improvements in packet delivery rates and energy efficiency of devices.