NoiseCo: Smartphone-based noise collection and correction

Nowadays, noise pollution has become an outstanding problem in modern life since it affects people's health physically and psychologically. However, the means for people to understand the noise level around them are very limited. They can only get to know their environment noise through facilities such as noise measuring station or noise display board in most cases. And the traditional ways to monitor and measure noise level by distributed networks and fixed points cannot meet the need of residents and research works anymore. So developing a new way to measure and monitor noise is very urgent. Mobile crowdsensing uses the power of the masses. It is cheap and very suitable as a way to measure noise on a large scale. By implementing a prototype called NoiseCo, this paper solves the accuracy problem of measuring noise by smartphones and proves the feasibility of monitoring noise level by mobile crowdsensing. Two major problems in measuring noise level by smartphones are how to calibrate the differences between different smartphones and how to calibrate the differences between smartphones and standard noise measuring tools. By experiments, we find out that the data collected by smartphones have a linear relationship with the data collected by the noise meter and we solve the problems according to this discovery. At the same time, we analyze the importance of sampling methods and A-weighting.