2005IEEE Transactions on Information TheoryRequires access

The Capacity of Finite-State Markov Channels With Feedback

Jun Chen, Thomas Berger

Open publisher page 133 citations

Abstract

We consider a class of finite-state Markov channels with feedback. We first introduce a simplified equivalent channel model, and then construct the optimal stationary and nonstationary input processes that maximize the long-term directed mutual information. Furthermore, we give a sufficient condition under which the channel's Shannon capacity can be achieved by a stationary input process. The corresponding converse coding theorem and direct coding theorem are proved.

About this research paper

What this paper is about

We consider a class of finite-state Markov channels with feedback. We first introduce a simplified equivalent channel model, and then construct the optimal stationary and nonstationary input processes that maximize the long-term directed mutual information. Furthermore, we give a sufficient condition under which the channel's Shannon capacity can be achieved by a stationary input process. The corresponding converse coding theorem and direct coding theorem are proved.

Why it matters

OpenAlex reports 133 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

We consider a class of finite-state Markov channels with feedback. We first introduce a simplified equivalent channel model, and then construct the optimal stationary and nonstationary input processes that maximize the long-term directed mutual information. Furthermore, we give a sufficient condition under which the channel's Shannon capacity can be achieved by a stationary input process. The corresponding converse coding theorem and direct coding theorem are proved.

Key concepts: Converse, Markov process, Markov chain, Channel capacity, Mathematics, Mutual information, Finite state, Coding (social sciences)

Related papers

Back to paper searchBrowse research topicsOriginal source
The Capacity of Finite-State Markov Channels With Feedback — Research Paper | ScholarLens