Design aspects of acoustic sensor networks for environmental noise monitoring

Abstract

An increase in public awareness of noise pollution and the impact of noise on human health has led to the need for enhanced insight in complex noise situations. This insight is commonly obtained either by brief measurements or by evaluation of a simplified acoustic model. Both of these approaches however have limitations in complex noise situations. Noise monitoring can be an appropriate and cost efficient measure to obtain more insight, for it allows to measure at many locations for long periods of time. Monitoring can for example be used to improve the accuracy of models or to assess and respond to changes in the acoustic situation. Several monitoring approaches, for various applications, have been developed or are under development. In the first part of this paper an overview is given of current developments in acoustic monitoring networks and their key aspects. Sensor networks for environmental noise monitoring are here divided into four different categories, distinguished by five aspects: hardware costs, scalability, flexibility, reliability and accuracy. These five aspects determine the range of applications for which a network is suited. In the second part of this paper a monitoring network developed in-house is used to further illustrate the relevance of these aspects. This network was designed to facilitate research into the field of acoustic monitoring networks and is used to experiment with and learn from a broad field of applications. Since the network is not designed for just one application, it has a special focus on flexibility and scalability. This allows the network to cope with the requirement of various applications with different scales. In the third section of this paper three applications of the network are shown to illustrate the five aspects. cop. 2016 Elsevier Ltd. All rights reserved.