English

Universal Covertness for Discrete Memoryless Sources

Information Theory 2021-06-21 v2 math.IT

Abstract

Consider a sequence XnX^n of length nn emitted by a Discrete Memoryless Source (DMS) with unknown distribution pXp_X. The objective is to construct a lossless source code that maps XnX^n to a sequence Y^m\widehat{Y}^m of length mm that is indistinguishable, in terms of Kullback-Leibler divergence, from a sequence emitted by another DMS with known distribution pYp_Y. The main result is the existence of a coding scheme that performs this task with an optimal ratio m/nm/n equal to H(X)/H(Y)H(X)/H(Y), the ratio of the Shannon entropies of the two distributions, as nn goes to infinity. The coding scheme overcomes the challenges created by the lack of knowledge about pXp_X by a type-based universal lossless source coding scheme that produces as output an almost uniformly distributed sequence, followed by another type-based coding scheme that jointly performs source resolvability and universal lossless source coding. The result recovers and extends previous results that either assume pXp_X or pYp_Y uniform, or pXp_X known. The price paid for these generalizations is the use of common randomness with vanishing rate, whose length scales as the logarithm of nn. By allowing common randomness larger than the logarithm of nn but still negligible compared to nn, a constructive low-complexity encoding and decoding counterpart to the main result is also provided for binary sources by means of polar codes.

Keywords

Cite

@article{arxiv.1808.05612,
  title  = {Universal Covertness for Discrete Memoryless Sources},
  author = {Remi A. Chou and Matthieu R. Bloch and Aylin Yener},
  journal= {arXiv preprint arXiv:1808.05612},
  year   = {2021}
}

Comments

11 pages, two-column, 2 figures, accepted to IEEE Transactions on Information Theory

R2 v1 2026-06-23T03:36:09.080Z