Skip to Main Content
Huffman coding is an efficient coding technique that is largely used in applications that need a lossless data compression. We present in this paper a novel method for Huffman decoding that increases the performance and keeps the memory consumption low. The main idea is to add information about the number of bits necessary to search for the next possible symbol in the Huffman tree. The proposed method has been used in the implementation of an MPEG-4 AAC decoder, and has proven to be extremely fast to decode short symbols that are the ones with higher probability to occur in a bitstream.
Date of Conference: 9-13 Jan. 2008