Dispersion Bound for the Wyner-Ahlswede-Körner Network via Reverse Hypercontractivity on Types

05/09/2018
by   Jingbo Liu, et al.
0

Using the functional-entropic duality and the reverse hypercontractivity of the transposition semigroup, we lower bound the error probability for each joint type in the Wyner-Ahlswede-Körner problem. Then by averaging the error probability over types, we lower bound the c-dispersion (which characterizes the second-order behavior of the weighted sum of the rates of the two compressors when a nonvanishing error probability is small) as the variance of the gradient of _P_U|X{cH(Y|U)+I(U;X)} with respect to Q_XY, the per-letter side information and source distribution. On the other hand, using the method of types we derive a new upper bound on the c-dispersion, which improves the existing upper bounds but has a gap to the aforementioned lower bound.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
11/08/2021

A Sphere-Packing Error Exponent for Mismatched Decoding

We derive a sphere-packing error exponent for coded transmission over di...
research
02/04/2018

A Sharp Bound on the s-Energy

We derive an (essentially) optimal bound of O(1/ρ s)^n-1 on the s-energy...
research
05/16/2018

Remote Source Coding under Gaussian Noise : Dueling Roles of Power and Entropy Power

The distributed remote source coding (so-called CEO) problem is studied ...
research
10/02/2014

A Lower Bound for the Optimization of Finite Sums

This paper presents a lower bound for optimizing a finite sum of n funct...
research
04/10/2017

Reinterpreting Importance-Weighted Autoencoders

The standard interpretation of importance-weighted autoencoders is that ...
research
10/23/2021

Signal to Noise Ratio Loss Function

This work proposes a new loss function targeting classification problems...
research
09/05/2022

Arithmetical Hierarchy of the Besicovitch-Stability of Noisy Tilings

The purpose of this article is to study the algorithmic complexity of th...

Please sign up or login with your details

Forgot password? Click here to reset