Efficient Huffman Decoding using Canonical Huffman Tree

We present an efficient decoding scheme for Huffman codes in which we use a properties of canonical prefix tree. After Huffman tree is converted to canonical Huffman tree, we represent Huffman tree with minimum information using rules associated with values of nodes in canonical tree. The proposed scheme can reduce memory to store Huffman tree information while maintains the same Processing time. The memory size in order to represent tree information is 2h + 2klogn which is less than those of previous methods. But the number of search is similar to previously proposed techniques.