Acoustic categorization of the urban multi-sensor network of the DYNAMAP LIFE project developed for road traffic noise mapping

The DYNAMAP LIFE project proposes the implementation of a dynamic noise mapping system able to determine the acoustic impact of road infrastructures in real-time, following the European Noise Directive 2002/49/EC. A Multi-Sensor Network collects the noise level measurements in two pilot areas: in the city of Milan and in the A90 motorway around Rome. For a proper evaluation of the equivalent noise level of road infrastructures, the anomalous noise events (ANE) unrelated to traffic noise (e.g. sirens, horns, speech, doors...) should be removed automatically. For this purpose, an anomalous noise events detector (ANED) has been designed using Mel-Frequency Cepstral Coefficients (MFCC) and Gaussian Mixture Models (GMM). This work focuses on the analysis of the spectral characteristics of the acoustic data from the 24-nodes of the Milan urban area network in order to determine whether a single adjustment of the ANED can cope with the particularities of each sensor location. To that effect, representative acoustic data has been collected through the Multi-Sensor network in real-operation conditions. These data have been subsequently analyzed in terms of their spectrotemporal distribution as a first step of the generalization of the ANED algorithm for all the nodes in the WASN.