Hidden Digital Modulation of Audio Signals

Abstract

Introduction. A modern methods of digital signal processing are capable to embed imperceptibly an additional information into audio signals using so called digital watermarks (DWM). DWM are characterized by the main parameters: payload, fidelity and robustness. One of the most important methods for DWM is based on quantization of certain signal parameters. But quantization methods are very sensitive to amplitude scaling and doesn’t proved appropriate fidelity and robustness.
The main part. The designed embedding algorithm is based on quantization of correlation coefficient between signal vector generally in a complex L-dimensional space and real random vector. Algorithm keeps signal norm that allows signal normali zation in modulation and demodulation processes and minimizes the embedding distortions. Variants of correlation coefficient quantization in the forms of Quadrature Amplitude Modulation and Amplitude Modulation in the frequency domain of discrete Fourier transform are considered.
The proposed multichannel algorithm provides robustness against intersymbol interference (ISI). It is proved advantages of data embedding in the frequency domain versus time domain and row-wise forming of signal vector for ISI resistance.nIt is shown by computer simulation the ability of flexible regulation of watermark performances by means of appropriate choosing of algorithm parameters. Practical implementation of the proposed algorithm is based on sliding processing in the whole and comparison of received and calculated hash functions.
Conclusion. The proposed algorithm demonstrates watermarks inaudibility because of preserving signal power before and after data embedding and proportional watermarking according to instantaneous carrier signal power. Watermark payload – fidelity – robustness compromise is provided by appropriate parameters choosing N, L, K, Δ in watermarking algorithm. The designed algorithms are applicable in mobile maritime and aeronautical radiotelephone services and resolving “analog hole” problem in the digital rights management for copy control of audio files.