Design and Simulation of Audio Watermarking using Empirical Mode Decomposition for Copyright Protection

Digital watermarking is drawing attention a new method for protecting multimedia content from illegal copying. In this paper, we present a alternative audio watermarking procedure to provide copyright protection to the digital audio by modifying audio samples directly by making use of algorithm called Empirical Mode Decomposition (EMD). The original audio signal is sampled at 44.1 kHz to split it into non-overlapping frames. These frames are further decomposed by EMD adaptively into intrinsic oscillatory mono-component signals known as Intrinsic Mode Functions (IMFs). The digital watermark along with synchronization codes are embedded into the last IMF extrema. This is low frequency mode, stable under different attacks and preserving the audio quality of host signal. Based on the simulations, robustness of hidden watermark is demonstrated for different attacks such as noise, filtering, compression, cropping, resampling and re-quantization.. It is seen that the data embedding rate or payload of the algorithm is 46.9-50.3 bits per second.