A Deep Learning-Powered FastAPI Web Application for Diabetes Mellitus Prediction:Model Development and Evaluation

Abstract

This project proposes an IoT-based dual power mobile network monitoringsystemdesigned to ensure continuous connectivity during power outages. It involvestwonetwork towers—one connected only to the grid and the other integrated withbothgrid and solar power sources. A physical switch is used to simulate load sheddingconditions. Upon load shedding, both towers initially drop from4Gto 2G. Thetowerequipped with solar power automatically recovers and restores the networkfrom2Gback to 4G using solar energy. The system includes a display that visuallyrepresents the current network status and power source. It also features anIoTplatform that updates the real-time condition of each tower, showing messages suchas "LOAD SHEDDING, Taking Power From Solar Panel" or "NO LOADSHEDDING, Taking Power From Grid." This project demonstrates an efficient approachto powerredundancy, network stability, and remote monitoring for future-ready telecominfrastructure