Coding of sources with two-sided geometric distributions and unknown parameters

Lossless compression is studied for a countably infinite alphabet source wi th an unknown, off-centered, two-sided geometric (TSG) distribution, which is a commonly used statistical model for image prediction residuals. In thi s correspondence, we demonstrate that arithmetic coding based on a simple s trategy of model adaptation, essentially attains the theoretical lower boun d to the universal coding redundancy associated with this model. We then fo cus on more practical codes for the TSG model, that operate on a symbol-by- symbol basis, and study the problem of adaptively selecting a code from a g iven discrete family. By taking advantage of the structure of the optimum H uffman tree for a known TSG distribution, which enables simple calculation of the codeword of every given source symbol, an efficient adaptive strateg y is derived.