Development of adaptive de-noising algorithm using wavelet technique for a linear FM acoustic signal

Abstract

Background/Objectives: It is highly important to analyze on the challenges that influence the acoustic medium. The underwater signals are affected by variety of background noise (ambient noise) in the ocean which is both natural and man-made. Method/Statistical Analysis: The methodology involved in this research work is wavelet decomposition technique to reduce the underwater noise present in the acoustic signal to extract the details present in it. Findings: In this research work, a detailed knowledge on the ambient noise which was collected from the shallow water region of Bay of Bengal, was obtained by characterizing it and a suitable denoising algorithm was formulated using wavelet technique, in particular Gabor wavelet and improvement in SNR is verified using MAT lab simulink tool. Applications/Improvements: The applications of underwater acoustics includes exploration of the environment, monitoring and tracking the marine mammals, exploration of oil in the deep ocean, monitoring the submarines and underwater autonomous vehicles etc,. A comparison was done between Gabor wavelet and Symlet wavelet in denoising the noisy acoustic signal to improve the signal to noise ratio. It enhances the performance of Gabor wavelet. From the result, we can understand that, for an input SNR range of -15 db to 0 db, we obtained an output SNR in the order of 9dB, 15db and 14db at 20 KHz, 66 KHz and 86 KHZ respectively which enhances the performance of Gabor wavelet.