Robust audio watermarking algorithm based on DWT using Fibonacci numbers

This paper presents a blind and robust audio watermarking algorithm developed based on Fibonacci numbers properties and the discrete wavelet transform (DWT) advantages. The method embeds watermark bits in the 6th level approximation subband of DWT at which there is less sensitivity of the human auditory system. The key idea is dividing the 6th level approximation coefficients into small frames and modifying their magnitude based on Fibonacci numbers and watermark bit values. The proposed watermarking method demonstrates a superior robustness against different common attacks (i.e., Gaussian noise addition, Low-pass filter, Resampling, Requantizing, MP3 compression, Amplitude scaling, Echo addition, Time shift, and Cropping). Compared to recently developed methods, the proposed algorithm is much more robust against the most common attacks with capacity as high as 686 bits per second. The results of PEAQ testing verify the quality of watermarked audio signal without significant perceptual distortion. The algorithm allows flexibility in audio watermark algorithm to achieve a balance between robustness and imperceptibility while the capacity is maintained constant by choosing various kinds of sequence.