Ferramenta de diagnóstico automatizado para monitoramento da qualidade em redes residenciais

Abstract

The accelerated growth in the number of connected devices in Brazilian households, driven by recent technological and social transformations, has exposed vulnerabilities in residential network infrastructure, particularly during the pandemic period. In this context, this dissertation proposed the development of an automated network diagnostic tool aimed at residential users without advanced technical knowledge. The central objective is to identify, monitor, and recommend corrective actions for both local and Internet connectivity issues. The theoretical framework addressed the principles of local area network (LAN) and wire area network (WAN) operation, IEEE 802.11 standards, quality of services (QoS) and experience (QoE) metrics, domain name server (DNS) tests, and bandwidth measurement tools. The implementation methodology consisted of developing a Python-based application that integrates continuous metric collection modules, statistical analysis, and the generation of interpretable diagnostics through an intuitive web interface. The developed tool performs segmented tests by network layer (LAN, wireless local area network (WLAN), and WAN), stores results in a not only SQL (NoSQL) database, and presents data in real-time, along with technical reports and historical graphs. The system also includes an automated mechanism, which classifies network status according to severity levels and recommends corrective measures. The solution was validated in a real residential environment, where it was possible to identify and mitigate connectivity failures through the tool’s recommendations. The results demonstrated the system’s effectiveness in proactively detecting performance degradation, interpreting technical metrics in an accessible manner, and empowering users to make autonomous decisions. It is concluded that the proposed objectives were fully achieved, with the delivery of a functional, scalable, and technically grounded solution that significantly contributes to the digital empowerment of residential users. This research also paves the way for future applications, such as integration with intelligent assistants, the use of machine learning for fault prediction, and the construction of collaborative databases for large-scale Internet quality assessment.