A cross spread spectrum audio watermarking robust to acoustic transmission

This paper proposes a technique to improve the robustness of spread spectrum (SS) audio watermarking for acoustic propagation. The appropriate embedding areas should be selected to achieve good perceptibility and high robustness. In order to improve the redundancy and robustness, the cross spread spectrum (CSS) scheme based on highly correlated cross frames is proposed which could decrease the variance of decision statistic and lead to a small error probability. Then we propose an improved cross spread spectrum (ICSS) scheme by incorporating CSS and improved spread spectrum (ISS). Our theoretical analysis shows that ICSS significantly reduces the host interference and improve the watermarking decoding performance. Simulation results demonstrated that the proposed ICSS algorithm is more robust not only to room reverberation and environmental noises, but also to MP3 compression and a variety of other distortions in Stirmark Benchmark for Audio.