English

Probability Mass Functions for which Sources have the Maximum Minimum Expected Length

Information Theory 2019-03-12 v1 math.IT

Abstract

Let Pn\mathcal{P}_n be the set of all probability mass functions (PMFs) (p1,p2,,pn)(p_1,p_2,\ldots,p_n) that satisfy pi>0p_i>0 for 1in1\leq i \leq n. Define the minimum expected length function LD:PnR\mathcal{L}_D :\mathcal{P}_n \rightarrow \mathbb{R} such that LD(P)\mathcal{L}_D (P) is the minimum expected length of a prefix code, formed out of an alphabet of size DD, for the discrete memoryless source having PP as its source distribution. It is well-known that the function LD\mathcal{L}_D attains its maximum value at the uniform distribution. Further, when nn is of the form DmD^m, with mm being a positive integer, PMFs other than the uniform distribution at which LD\mathcal{L}_D attains its maximum value are known. However, a complete characterization of all such PMFs at which the minimum expected length function attains its maximum value has not been done so far. This is done in this paper.

Cite

@article{arxiv.1903.03755,
  title  = {Probability Mass Functions for which Sources have the Maximum Minimum Expected Length},
  author = {Shivkumar K. Manickam},
  journal= {arXiv preprint arXiv:1903.03755},
  year   = {2019}
}
R2 v1 2026-06-23T08:02:55.774Z