19-10-2010, 05:46 PM
This article describes a efficient hardware architecture for Lempel-Ziv-Welch (LZW) data compression algorithm. The encoding and decoding operations are done simultaneously using a CAM array. In order to to achieve search and twofold store operations in single access during regular match operations, an anhanced CAM architecture is also proposed. the implementation of the LZW algorithm is accelerated by these enhanced CAM cells. the proposed
design is evaluated using the Corpus benchmarks.
Introduction
many lossless data compression algorithms such as LempelZiv (LZ) Huffman coding and arithmetic coding are in use today. In LZW-based data compression the dictionary is constructed dynamically without any prior knowledge of source data. GIF, TIFF, ZIP and postscript level 2 etc uses the LZW algorithm.
The sequential construction of a dictionary is an essential part in the LZW algorithm and then the compression technique involves the involves extensive comparison of input data with the dictionary content.
LZW Compression
First, the dictionary table is initialized to contain all single character strings. the code is stored as prefix string if a match is found. After this, the next character is read and the the substring consisting of prefix string and the next character. If
there is a mismatch the prefix string is sent to the output
and the input character (K) becomes the first character
of the next string. If it matches, the string is then assigned to a prefix string. until all the inputs are exhausted, the characters are read.
Get the report here:
http://mediafire?5ampukicfbdgbsu
