一种数据无损压缩技术的研究

A Study for Nondestructive Compression of Data

分析了目前数据压缩技术中最常用的ＬＺＷ算法，并针对压缩字典的建立过程和字典填满老化后如何更新处理提出了新的设想。在实现过程中，用原ＬＺＷ算法中新加入字典的词条作为辅助前缀来产生较多较长的词条，用以加速压缩字典的建立过程，增加从字典中找到词条匹配的概率，在字典填满并老化后，采用部分更新字典的办法加速字典的重建，从而对原ＬＺＷ算法作了两点改进。本文举例说明了两种算法的不同之处，并经实验比较证明，改进后的算法在数据压缩率方面优于原ＬＺＷ算法。

Analyses the LZW algorithm common used in the area of lossless data compression technique. A new idea is presented to handle the creation of compressed dictionary and to renew the dictionary which is filled up and olden. In the realization, many longer words are derived by taking some new words added into the dictionary in the ordainal LZW algorithm as supplementary prefixes so as to fasten the creation of compressed dictionary and to raise the probability of finding matched-words in the dictionary. When the dictionary is full and olden, the reconstruction of dictionary is acCelerated by renewing it partly. Two points of approvement have been made for LZW algorithm. There is an example in this paper used to illustrate the difference of two algorithms. The experiment proves that the improved algorithm is better than LZW one in lossless compression of data.