Detecção de incêndios florestais em áreas remotas através de redes LoRa

• Ano de defesa: 2024
• Autor(a) principal: Araújo, Mateus Sousa
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Incêndios florestais; Veículos aéreos não tripulados; Diversidade espacial; Simulação; Agrupamento de gateways; CNPQ: CIENCIAS EXATAS E DA TERRA: CIENCIA DA COMPUTACAO
• Link de acesso: http://repositorio.ufc.br/handle/riufc/77707

Wildfires are a critical problem, especially in remote areas, where rapid detection and response are essential to minimize environmental and economic damage. The need for efficient early detection systems is vital to ensure that alerts are issued accurately and that firefighting resources are mobilized quickly. In this scenario, LoRa technology emerges as a promising solution for wildfire detection due to its long-distance communication capability, low power consumption, and reduced cost. This monograph addresses the problem of reliably detecting wildfires, followed by delivering warning signals in an efficient and timely manner. Establishing a sensor network in a remote and inhospitable location is, in itself, a great challenge. This work implements a wildfire detection system assisted by some unmanned aerial vehicles (UAVs) capable of scanning a large area. The results were obtained using a proprietary computational simulation tool that allows integrating LoRa communication models with the spatial diversity inherent to UAVs and smoke dispersion models. The performance evaluation of this system was carried out based on open data from the Altamira conservation unit. The results obtained through the simulation demonstrated a significant improvement in the delivery of LoRa packets, using gateway clustering techniques and application of spatial diversity in the network elements. Extensive simulation campaigns have shown that the combination of these techniques improved the reliability of fire alerts, even in adverse conditions, thus providing efficient detection of wildfires in remote areas.