Stochastic Interpretation for the Arimoto Algorithm
Abstract
The Arimoto algorithm computes the Gallager function for a given channel and parameter , by means of alternating maximization. Along the way, it generates a sequence of input distributions , , ... , that converges to the maximizing input . We propose a stochastic interpretation for the Arimoto algorithm. We show that for a random (i.i.d.) codebook with a distribution , the next distribution in the Arimoto algorithm is equal to the type () of the feasible transmitted codeword that maximizes the conditional Gallager exponent (conditioned on a specific transmitted codeword type ). This interpretation is a first step toward finding a stochastic mechanism for on-line channel input adaptation.
Cite
@article{arxiv.1412.4510,
title = {Stochastic Interpretation for the Arimoto Algorithm},
author = {Sergey Tridenski and Ram Zamir},
journal= {arXiv preprint arXiv:1412.4510},
year = {2015}
}
Comments
5 pages, 1 figure, accepted for 2015 IEEE Information Theory Workshop, Jerusalem, Israel