Computation of Marton's Error Exponent for Discrete Memoryless Sources

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The error exponent of fixed-length lossy source coding was established by Marton. Ahlswede showed that this exponent can be discontinuous at a rate R, depending on the probability distribution P of the given information source and the distortion measure d(x, y). The reason for the discontinuity in the error exponent is that there exists (d, ) such that the rate-distortion function R(|P) is neither concave nor quasi-concave with respect to P. Arimoto's algorithm for computing the error exponent in lossy source coding is based on Blahut's parametric representation of the error exponent. However, Blahut's parametric representation is a lower convex envelope of Marton's exponent, and the two do not generally agree. The contribution of this paper is to provide a parametric representation that perfectly matches the inverse function of Marton's exponent, thus avoiding the problem of the rate-distortion function being nonconvex with respect to P. The optimal distribution for fixed parameters can be obtained using Arimoto's algorithm. Performing a nonconvex optimization over the parameters successfully yields the inverse function of Marton's exponent.

Original languageEnglish
Title of host publication2023 IEEE International Symposium on Information Theory, ISIT 2023
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1372-1377
Number of pages6
ISBN (Electronic)9781665475549
DOIs
Publication statusPublished - 2023
Externally publishedYes
Event2023 IEEE International Symposium on Information Theory, ISIT 2023 - Taipei, Taiwan, Province of China
Duration: Jun 25 2023Jun 30 2023

Publication series

NameIEEE International Symposium on Information Theory - Proceedings
Volume2023-June
ISSN (Print)2157-8095

Conference

Conference2023 IEEE International Symposium on Information Theory, ISIT 2023
Country/TerritoryTaiwan, Province of China
CityTaipei
Period6/25/236/30/23

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Information Systems
  • Modelling and Simulation
  • Applied Mathematics

Fingerprint

Dive into the research topics of 'Computation of Marton's Error Exponent for Discrete Memoryless Sources'. Together they form a unique fingerprint.

Cite this