EDA-78: A Novel Error Detection Algorithm for Lempel-Ziv-78 Compressed Data

Beom Kwon, Myongsik Gong, Sanghoon Lee

Abstract: This paper presents an error detection algorithm for Lempel-Ziv-78 (LZ78) compressed data. LZ78 data compression involves dictionary coding and aims to compress the original data without loss. For detecting bit errors in compressed data, methods such as parity bit and Hamming code approaches have been applied. However, in these conventional methods, the insertion of additional bits is required for error detection, increasing the data redundancy. For error detection of LZ78 compressed data, we introduced four unique properties of LZ78 compressed data and developed an algorithm that detects bit errors in LZ78 compressed data according to these properties, without the insertion of additional bits. The proposed algorithm, which is called EDA-78 (Error Detection Algorithm for LZ78 compressed data), achieved an error detection rate of nearly 100% in the case of six or more bit errors. However, when the number of bit errors was smaller than six, the error detection rate was degraded. To overcome this drawback, we employed parity bits, significantly improving the error detection rate for a small number of bits.