展頻隱藏性聲音通訊之二層強化及其應用

Abstract

由於手持式運算設備的普及，隱藏性聲音通道是一個低成本的資訊傳遞工具。過去，此類通訊系統有兩派主流作法，其一利用正交分頻多工技術，傳輸效果好，但聲音品質差；其二利用展頻浮水印的技巧，將資訊隱藏在聲音當中，此法聲音品質較佳，但傳輸能力有限。 我們重新測量聲音通道的表現及特徵，進而針對展頻浮水印之通訊系統提出了兩個不同層次的強化方法：通訊方法階段採用了適應分頻技術，在不影響聲音品質的前提下，增加了傳輸能力；而在資料處理階段，利用霍夫曼編碼提昇傳輸效率。最終在整體表現上，我們維持了好的聲音品質，並達到了較好的傳輸速率。

Due to the growing popularity of handheld devices, using sound to pass information from audio source to mobile becomes a possible application. Although the throughput offered by channel is relatively low and the quality of audio is not as good as the original, it provides a costless solution for communication with mobile devices through acoustic channel. Many interesting applications based on such communication systems have been proposed. To reach the goal, spread spectrum watermarking technologies have been considered before, but the achieved throughput is very low. Another approach called “Acoustic OFDM” is proposed. The achieved transmission rate is superior to that of the SS-based method, but its associated quality of audio is controversial. We proposed two-level enhancement for spread spectrum watermarking based communication over the hidden aerial acoustic channel. It makes the throughput more than “Acoustic OFDM” method, but keeps the quality of audio as good as SS-based method.