Open AccessMulti-symbol Accessing Huffman Decoding Method for MPEG-2 AAC
Author(s) -
EunSeo Lee,
Kyoung-Cheol Lee,
Kyou-Jung Son,
Seong-Pil Moon,
TaeGyu Chang
Publication year - 2014
Publication title -
journal of electrical engineering and technology/journal of electrical engineering and technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.226
H-Index - 27
eISSN - 2093-7423
pISSN - 1975-0102
DOI - 10.5370/jeet.2014.9.4.1411
Subject(s) - huffman coding , code word , canonical huffman code , decoding methods , computer science , algorithm , arithmetic , computational complexity theory , symbol (formal) , table (database) , parallel computing , theoretical computer science , computer hardware , data compression , mathematics , code rate , programming language , data mining , systematic code
An MPEG-2 AAC Huffman decoding method based on the fixed length compacted codeword tables, where each codeword can contain multiple number of Huffman codes, was proposed. The proposed method enhances the searching efficiency by finding multiple symbols in a single search, i.e., a direct memory reading of the compacted codeword table. The memory usage is significantly saved by separately handling the Huffman codes that exceed the length of the compacted codewords. The trade-off relation between the computational complexity and the amount of memory usage was analytically derived to find the proper codeword length of the compacted codewords for the design of MPEG-2 AAC decoder. To validate the proposed algorithm, its performance was experimentally evaluated with an implemented MPEG-2 AAC decoder. The results showed that the computational complexity of the proposed method is reduced to 54% of that of the most up-to-date method.