A Universal 4D Model for Double-Efficient Lossless Data Compressions

Abstract

This article discusses the theory, model, implementation and performance of a combinatorial fuzzy-binary and-or (FBAR) algorithm for lossless data compression (LDC) and decompression (LDD) on 8-bit characters. A combinatorial pairwise flags is utilized as new zero/nonzero, impure/pure bit-pair operators, where their combination forms a 4D hypercube to compress a sequence of bytes. The compressed sequence is stored in a grid file of constant size. Decompression is by using a fixed size translation table (TT) to access the grid file during I/O data conversions. Compared to other LDC algorithms, double-efficient (DE) entropies denoting 50% compressions with reasonable bitrates were observed. Double-extending the usage of the TT component in code, exhibits a Universal Predictability via its negative growth of entropy for LDCs > 87.5% compression, quite significant for scaling databases and network communications. This algorithm is novel in encryption, binary, fuzzy and information-theoretic methods such as probability. Therefore, information theorists, computer scientists and engineers may find the algorithm useful for its logic and applications.

arXiv commentMajor changes have been made in this revision for a thorough and lucid representation of the concept following implementation, in which, abstract portions of it are considered for publication in Information Theory journals as well as a foundation work for a PhD dissertation at http://web.uvic.ca/~phibal12/publication.htm . Written in LaTeX: 76 pages, 56 references, 9 figures and 6 tables