Design and Implementation of a Smart Weather Station with Intelligent Communication Interface Selection for Reliable Data Transmission

Abstract

This paper presents the design and implementation of a smart weather monitoring system featuring an adaptive
network interface selection algorithm for reliable environmental data transmission. The system combines IoTbased sensing with intelligent communication switching logic that automatically selects between Wi-Fi and GSM
network interfaces base on connectivity strength and availability. This adaptive approach ensures uninterrupted
operation and efficient data delivery even under unstable network conditions. Each station integrates sensors for
temperature, humidity, pressure, rainfall, wind parameters, and UV index, managed by an ESP32-based
controller. A cloud-hosted API service built on FastAPI handles secure data reception, while a web dashboard
provides real-time visualization and historical data analysis. The firmware includes automatic configuration,
selective sensor power control, and deep-sleep cycles to optimize power consumption. Experimental results show
that the adaptive interface selection improves transmission reliability, reduces data loss, and maintains consistent
latency across variable network conditions, ensuring up to 100% transmission success rate within the simulated
network conditions. The system thus establishes a robust and scalable model for autonomous, self-optimizing
weather stations, particularly suited for deployment in areas with inconsistent network coverage.