Abstract
A network of low-power wireless sensors for the detection of air quality pollutants (acetone, toluene, ammonia, formaldehyde, ethanol, benzene, dichloromethane, acetic acid, xylene and dimethylacetamide) is described in this paper. The network consists of several sensor nodes responsible for receiving information and wirelessly, through ZigBee protocol, sending it to a gateway connected to the Internet. A cloud infrastructure is used to store the data of each node and to perform the data processing. Several autonomous nodes equipped with four different gas sensors are used to measure the pollutants concentration in air. Each node also includes a pump and a 3-way electric valve to carry the air samples to the sensor cell. The 3-way electric valve can be operated in order to switch between reference clean air or gas samples from the vials containing the samples. The communication between nodes and the gateway is based on the standard IEEE 802.15.4 (Zig-Bee protocol). Data processing and analysis have been performed with the use of principal component analysis (PCA) and artificial neural networks.
