Urban Air Pollution Monitoring by Neural Networks and Wireless Sensor Networks Based on LoRa

The increase in air pollutant emissions is a current concern. Due to this, the present work shows a network of sensor nodes sending information by LoRa protocol to the monitoring of emissions of harmful gases for health in urban environments. To do this, an electronic scheme is proposed for data acquisition with a smoothing of the signal from each sensor for noise elimination. Subsequently, data analysis is performed using an artificial neural network with the main objective of classifying the state of the air. As relevant results, the classification performance of 95% in tests and 90% in real conditions with the presentation of this information in real-time is obtained.