According to a research by the Swiss technology company IQAir, India's cities made up 12 of the 15 cities with the highest pollution levels and the country's cities had the eighth-most contaminated air overall. India's average PM2.5 concentration, weighted by population, was 53.3 micrograms per cubic metre (g/m3) in 2022 compared to the WHO's suggested yearly recommendation level of 5 g/m3. IQAir's World Air Quality Report uses PM2.5, or particulate matter with a diameter of less than 2.5 microns, as the primary air quality measure. Additionally, the average determined by IQAir exceeded the permissive level of 40 g/m3 established by India's Central Pollution Control Board. The wireless sensor networks are becoming an active research field because of their wide application. This research study shall focus on developing an air pollution forecasting approach using the DeepAR, an autoregressive recurrent network using the data collected from a wireless sensor network. DeepAR has a high prediction ability as the working principle is combined with deep learning and conventional probabilistic prediction.

PM2.5 concentration, air quality, DeepAR, autoregressive recurrent network, wireless sensor network