整合前向錯誤控制於多重敘述語音播放排程設計之研究

Abstract

本論文所提出的多重敘述語音傳輸系統，是將音框編碼輸出的參數以兩個封包分別傳送，以期利用互相獨立的傳輸路徑提升其音框播放品質。為了準確反應多重敘述傳輸的音框回復品質，我們根據E模型提出一新的音質預測模型，其主要概念是區隔單一或雙重封包的接收狀態並加以適度整合。進一步以此音質預測模型作為多重敘述傳輸系統的發展平台，整合設計其播放排程與前向錯誤控制，以期依網路環境的時變特性適應性地調整其系統參數。系統模擬証實，新的音質預測模型更為接近主觀聽覺測試結果，我們提出的多重敘述傳輸系統能有效對抗封包漏失並輸出最佳音質。

Packet loss and delay are two essential problems to real-time voice transmission over IP networks. In the proposed system, multiple descriptions of the speech are transmitted to take advantage of largely uncorrelated delay and loss characteristics on different network paths. Adaptive joint playout buffer and FEC adjustment of multiple voice streams is formulated as an optimization problem leading to a better delay-loss tradeoff. The basic strategy is a perceptually motivated optimization criterion based on a modified ITU-T E-model for multiple-stream transmission . Experimental results show that the proposed multi-stream voice transmission system improves the delay-loss tradeoff as well as speech reconstruction quality.