Block sorting transformations

Summary form only given. The Block Sorting Lossless Data Compression Algorithm (BSLDCA) described by Burrows and Wheeler (1994) has received considerable attention. BSLDCA achieves compression ratio closer to PPM, but with a faster execution speed than PPM. Arnavut and Magliveras (1997) described the theoretical basis for block sorting schemes in the case of permutations and introduced the Lexical Permutation Sorting Algorithm (LPSA). This paper, generalizes the theoretical foundations of block sorting schemes to the multiset permutations (data strings). By expanding the theoretical foundations of LPSA from permutations to multiset permutations (data strings), we introduce a different block transformation, linear order transformation (LOT), and delineate its relationship to the Burrows Wheeler transformation (BWT). We show that LOT is faster than BWT, and for certain data types, such as pseudo-color images, LOT transformation followed by the MTF coder, yields the same amount of compression.