Universal coding for arbitrarily varying sources

Meir Feder*, Neri Merhav

*Corresponding author for this work

Research output: Contribution to conferencePaperpeer-review

Abstract

The minimum universal coding redundancy for finite-state arbitrarily varying sources, is investigated. If the space of all possible underlying state sequences is partitioned into types, then this minimum can be essentially lower bounded by the sum of two terms. The first is the minimum redundancy within the type class and the second is the minimum redundancy associated with a class of sources that can be thought of as 'representatives' of the different types. While the first term is attributed to the cost of uncertainty within the type, the second term corresponds to the type itself. The bound is achievable by a Shannon code w.r.t. an appropriate two-stage mixture of all arbitrarily varying sources in the class.

Original languageEnglish
Pages16
Number of pages1
StatePublished - 1995
EventProceedings of the 1995 IEEE International Symposium on Information Theory - Whistler, BC, Can
Duration: 17 Sep 199522 Sep 1995

Conference

ConferenceProceedings of the 1995 IEEE International Symposium on Information Theory
CityWhistler, BC, Can
Period17/09/9522/09/95

Fingerprint

Dive into the research topics of 'Universal coding for arbitrarily varying sources'. Together they form a unique fingerprint.

Cite this