A case study on monitoring and geolocation of noise in urban environments using the internet of things

This paper presents the development of a system to monitor and geo-reference noise in urban environments using the Internet-of-Things (IoT). The system intends to help control agencies and citizens to monitor noise using smart devices and services available on the Cloud for data sharing. The system includes a mobile application that periodically captures the microphone's audio signal during a configurable time window, obtains the mobile's global position after each measurement using the built-in GPS, and assigns a times-tamp from the operating system. Then, a Fast Fourier Transform is applied to recorded audio and the power spectrum in decibels is extracted. The resulting vector is sampled at specific frequencies to create a vector of audio features that can be used to assess noise pollution completely offline. If an Internet connection is available, a telemetry message is assembled and sent to an IoT Hub in Microsoft Azure Cloud containing a unique ID, a position stamp, a time stamp, and all audio features. The message is transferred to a Stream Analytics service, and from there it is sent to a Cloud SQL Database for permanent storage. The historical information collected and shared by different users can be examined online by any individual through a customized report developed in Microsoft Power BI.