Unequal Error Correction for LZ77 Compressed Data
Tang Hong · Journal of Sichuan University · 2002
This paper proposed an efficient error control coding method for the data compressed by widely used LZ77 algorithm. Based on the analysis and computer simulation results of affections of errors in different components of compressed data, a Unequal Error Protection (UEP) coding method is recommended for the matched lengths, the offsets and unmatched characters of LZ77 compressed data. In general, matched lengths are better protected from errors than the other two because errors in matched lengths have much more serious affection to the decompressed data than the other two; and since the former the errors in matched length, the more serious the affections to decompressed data, while errors in other two have equal affections regardless their locations in compressed data, matched lengths are coded by using UEP codes, which can correct more errors in the former part. For the offsets and unmatched characters, an equal protection code is used. The principle of error control coding for LZ77 compressed data is that errors in matched length should be corrected as much as possible. Computer simulations show that the proposed method can control errors more efficiently in decompressed data than conventional error control codes although the redundancy introduced by proposed method is less than that of the conventional method.