Squeezing the Arimoto-Blahut algorithm for faster convergence

TL;DR

This paper introduces a 'squeezing' strategy to improve the convergence rate of the Arimoto-Blahut algorithm for channel capacity computation, maintaining its simplicity and convergence guarantees.

Contribution

The paper proposes a novel squeezing approach that accelerates the Arimoto-Blahut algorithm while preserving its key properties.

Findings

Faster convergence rates demonstrated theoretically.

Preservation of algorithm simplicity and monotonic convergence.

Applicable to discrete memoryless channel capacity calculations.

Abstract

The Arimoto--Blahut algorithm for computing the capacity of a discrete memoryless channel is revisited. A so-called ``squeezing'' strategy is used to design algorithms that preserve its simplicity and monotonic convergence properties, but have provably better rates of convergence.