Improved Algorithm LZW Compression Dictionary And FPGA Hardware

Along with the development of computer technology, data compression technology is bound to be a key technology for massive data storage and transmission. Data compression technology is mainly divided into lossless data compression and lossy data compression. The lossless data compression has been widely used in aerospace, medical and the other technical fields. The commonly used data lossless compression technology includes the Huffman code, the arithmetic code, traveling schedule code and so on. LZW compression is a dictionary-based lossless data compression method. While the statistical properties are unknown, the LZW dictionary compression has high adaptive and coding efficiency. LZW data compression algorithm has become a standard computer file compression algorithm and is widely used in image compression and texts compression.The main content of this article includes the following aspects. Firstly, the paper describes several common lossless data compression algorithms and related theoretical basis for data compression, compares several types of compression algorithms. Secondly, takes LZW data compression algorithm as the main research object, researches the encoded and decoded mode of LZW data compression algorithm and carries on the software simulation through MATLAB software to realize this algorithm. At the same time, an improved LZW data compression method is proposed in this paper based on original LZW compression insufficiencies. The LZW compression code mainly includes three insufficiencies, the first is that compression ratio is low in initial period of file compression, second is code output steam with fixed length, and the last is dictionary no longer to be updated after overflow. The new LZW compression algorithm mainly makes the improvement to these three aspects, which includes combining LZW and Huffman algorithm,8bit-12bit variable length output code steam, and FIFO dictionary update method. Thirdly, the improved LZW data compression is applied to English file compression and digital gray-scale image compression. Besides, the improved method is compared to original LZW compression algorithm and Huffman code, the advantages of new method is analyzed. Finally, the improved LZW data compression is carried on FPGA and each sub-module program is introduced in paper. According to the FPGA parallel processing data characteristic, the dictionary with4096capacity is divided into ten different small dictionaries. The dictionary width is variable, thus reduces the storage space and the chip area. At the same time the hardware system based on segmentation method using a parallel dictionary matches the search strategy, greatly improves the speed of string matching and coding speed.After the design completes, the simulation and the synthesis of the improved method is carried on platform of QuartusII. The feasibility and practicality of the LZW algorithm is verified.