雙向性可變長度碼的軟性訊源解碼機制

Abstract

近年來多媒體通訊的快速發展，使得針對影音資料的壓縮已成為重要的研究課題，特別是多項國際標準制定的可變長度碼。本論文主要探討可變長度碼的軟性訊源解碼機制，並將其應用於心電圖訊號的壓縮。首先針對可變長度碼在索引層級的籬柵結構，推導其軟性訊源解碼演算法，並整合其軟性輸出與籬柵結構以發展基於維特比演算法的索引估測方法。其次，針對具有特定機率分佈的訊源，可建構其霍夫曼碼，並以此發展具有錯誤更正能力的雙向性可變長度碼。我們將引入訊源解碼的軟性資訊，配合一階相關性的事前資訊進行索引指定的設計。系統模擬證實我們提出的方法能有效對抗通道雜訊干擾，進而提升系統的強健性。

Reversible variable length codes (RVLCs) were proposed for facilita- ting the bidirectional decoding, which mitigate the transmission errors in case of losing synchronization between the encode and decoder. This thesis focuses on the RVLC design and its application to ECG signal compression. Our work starts with a soft-decision source decoding algorithm based on a symbol-level trellis also proposed is a viterbi algorithm which exploit the soft output of source decoder to better estimate the transmitted index. For construction of RVLC, we proposed an index assignment scheme based on the channel-related information and the first-order correlation of source. Simulation results show that the proposed scheme can generate RVLC of higher robustness in comparison to other construct algorithms.