Design of IP Core for Sound Classification in Wireless Sensor Networks

This paper shows the current stage of development of an IP Core for sound classification in wireless sensor networks using Mel-frequency cepstrum (MFCC) and support vector machine (SVM) algorithms. This way a sensor node can do the job by itself instead of sensing to another node or a computer, this represents a great issue for power saving. The IP Core implementation was validated in software doing bird identification.