Banca de DEFESA: GABRIEL MAIA DE MACEDO

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : GABRIEL MAIA DE MACEDO
DATE: 23/07/2026
TIME: 09:00
LOCAL: Link da videochamada: https://meet.google.com/rhi-mbns-gpr
TITLE:

A Machine Learning-Based Rate Adaptation Mechanism for LoRaWAN Networks in Semi-Intensive Livestock Farming

KEY WORDS:

LoRaWAN, LoRa, ADR, Livestock, Machine Learning, Reinforcement Learning, IoT.

PAGES: 83
BIG AREA: Engenharias
AREA: Engenharia Elétrica
SUBÁREA: Telecomunicações
SPECIALTY: Sistemas de Telecomunicações
SUMMARY:

This work aims to explore Adaptive Data Rate (ADR) techniques in LoRaWAN networks for cattle monitoring in semi-intensive livestock systems, evaluating network performance and energy consumption based on the number of monitored animals and coverage area. The proposed approach was evaluated through simulations using ns-3. The simulation scenario considers the seasonal movement of animals, alternating between a confinement period in a corral and a grazing period in the field. An analysis of classical ADR techniques applied to this scenario shows that different ADR algorithms perform better at specific moments. Based on this evidence, this work employs reinforcement learning techniques to dynamically adjust transmission parameters, adapting to the dynamic nature of the proposed scenario. The results show that the adoption of machine learning leads to significant network gains in terms of packet delivery reliability, data rate at the Gateway, and energy efficiency of the End devices.

COMMITTEE MEMBERS:
Presidente - 1412682 - VICENTE ANGELO DE SOUSA JUNIOR
Interno - 2524053 - ANTONIO LUIZ PEREIRA DE SIQUEIRA CAMPOS
Externo à Instituição - Waslon Terllizzie Araujo Lopes. - UFPB
Externo à Instituição - ÁLVARO AUGUSTO MACHADO DE MEDEIROS - UFJF