Analysis and Practice of Uniquely Decodable One-to-One Code

Abstract

In this paper, we consider the uniquely decodable one-to-one code (UDOOC) that is obtained by inserting a comma indicator, termed the unique word (UW), between consecutive one-to-one codewords for separation. As such, we analyze a class of UDOOCs and present practical algorithms for encoding and decoding such codes. Specifically, for various cases of UWs, we investigate the number of length-n codewords of UDOOCs and their asymptotic growth rates in n. The proposed encoding and decoding algorithms of UDOOCs can be implemented in parallel at low computational complexity without storing the codebook. Simulation results show that for proper choices of UWs, UDOOCs can achieve better compression efficiency than Lempel-Ziv codes even when the source is not statistically independent.