Balancing Robustness and Inaudibility in Audio Watermarking: A Wavelet Component Proportional Coefficient Approach

Abstract

Audio watermarking research has predominantly prioritized robustness, yet most existing algorithms fail to achieve high robustness while maintaining complete inaudibility. To address this trade‑off, this paper proposes a digital watermarking algorithm based on wavelet component proportional coefficients. The algorithm first computes wavelet coefficients via Discrete Wavelet Transformation (DWT), then derives the wavelet coefficient proportions of two sub‑segments within each frame, and subsequently generates the wavelet coefficients of all sub‑segments through two‑modular addition. Finally, the watermark is embedded via inverse DWT based on the wavelet coefficients scaled by the coefficient proportions obtained before and after the process. Simulation results demonstrate that the proposed algorithm achieves favorable inaudibility and robust resistance to conventional signal processing operations, including MP3 compression, low‑pass filtering, and resampling.