Acoustic Signal Classification Algorithm for WSN Node in Transport System

In the paper, we focus on the classification of the acoustic signal and its characteristic properties, which we use for further processing of the acoustic signal. Its further processing is ensured that we are able to find the carrier frequencies of the selected signal with frequency analysis. We use compression methods to reduce the data needed to classify acoustic signals. We use neural networks to classify these signals. In addition, a method has been proposed to classify acoustic signals that are commonly found in transport. The result is the design of a method that is able to classify signals characteristic for different environments or different acoustic sources. In the paper, there is a description of the experiment that has been carried out for the mentioned purposes. For experiment is created evaluation and classification success rate on selected acoustic signals.

[1]  Michal Kochlán,et al.  Universal synchronization algorithm for wireless sensor networks—“FUSA algorithm” , 2014, 2014 Federated Conference on Computer Science and Information Systems.

[2]  Roman Neruda,et al.  Teoretické otázky neuronových sítí , 1996 .

[3]  Ondrej Karpis,et al.  Water level monitoring based on the acoustic signal using the neural network , 2016, 2016 International Conference on Information and Digital Technologies (IDT).

[4]  Juraj Micek,et al.  Monitoring of water level based on acoustic emissions , 2015, 2015 IEEE Symposium on Computers and Communication (ISCC).

[5]  Michal Hodon,et al.  Monitoring and recognition of bird population in protected bird territory , 2015, 2015 IEEE Symposium on Computers and Communication (ISCC).

[6]  S. Astapov,et al.  A method of real-time mobile vehicle identification by means of acoustic noise analysis implemented on an embedded device , 2012, 2012 13th Biennial Baltic Electronics Conference.