Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

Optica Advanced Photonics Congress

Neural Network Based Black-box Estimator of Mutual Information

Tobias A. Eriksson, Diego Villafani, Henrik Åhlfeldt, Laurent Schmalen

Published Jan 1, 2022• 0 citations
Publisher

Abstract

We show that neural networks with softmax outputs are capable of estimating a lower bound on the mutual information of memoryless communication channels without prior knowledge of the channel properties.
DOI
10.1364/sppcom.2022.sptu1j.3

Related papers

Study on estimating quantum discord by neural network with prior knowledge

Yong-Lei Liu, An Min Wang, Yi Sun +2 more

arXiv:1905.10782•Quantum Information Processing•Oct 10, 2019•4 citations

Four-dimensional estimates of mutual information in coherent optical communication experiments

Tobias A. Eriksson, Tobias Fehenberger, Norbert Hanik +3 more

Mutual information is experimentally investigated for long-haul coherent transmission. Receivers that consider memoryless four-dimensional noise distributions can achieve significantly higher rates than receivers assuming two-dimensional symmetric distributions.

European Conference on Optical Communication•Sep 1, 2015•6 citations

Exploring the Interpretability of Quantum Neural Networks by Mutual Information

Xin Zhang, Xiaoyu Li, Chenguang Zhang +1 more

•Jan 1, 2025•0 citations

Neural Estimation of Information Leakage for Secure Communication System Design

Darius S. Heerklotz, Ingo Schröder, Pin-Hsun Lin +2 more

Underestimating the leakage can compromise physical layer security (PLS), while overestimating it may lead to inefficient system design. Therefore, a reliable leakage estimator is essential. Neural network-based estimators provide a data-driven way to estimate mutual information (M.I.) without requiring full knowledge of the channel or source distributions. In this work, we aim to scale the blocklength of a wiretap code such that the estimator can still feasibly operate. We propose an improved M.I. estimator based on the variational contrastive log-ratio upper bound (vCLUB) framework, tailored for both discrete and continuous variables. By using a mixture of Bernoulli experts parameterized by neural networks, the estimator is able to quantify information leakage in communication systems, which employ complex data processing like universal hash families (UHF). We further propose a method to utilize the proposed estimator to design the UHF for a wiretap code or secret key generation design. Simulation results show that prior methods significantly underestimate the M.I., particularly when using UHF for higher blocklengths (n ≫16). The proposed method can scale the blocklength up to 255, and we conjecture that the design can scale well to even higher blocklengths given adequate training data and model size. Additionally, we contend that our proposed estimator and adaptive hash design framework offer a practical approach for extending PLS considerations for wiretap channels into the finite blocklength regime.

arXiv:2508.05176•IEEE Globecom Workshops (GC Wkshps)•Dec 8, 2025•4 citations

One-Shot Entanglement-Assisted Quantum and Classical Communication

Nilanjana Datta, Min-Hsiu Hsieh

We study entanglement-assisted quantum and classical communication over a single use of a quantum channel, which itself can correspond to a finite number of uses of a channel with arbitrarily correlated noise. We obtain characterizations of the corresponding one-shot capacities by establishing upper and lower bounds on them in terms of the difference of two smoothed entropic quantities. In the case of a memoryless channel, the upper and lower bounds converge to the known single-letter formulas for the corresponding capacities, in the limit of asymptotically many uses of it. Our results imply that the difference of two smoothed entropic quantities characterizing the one-shot entanglement-assisted capacities serves as a one-shot analog of the mutual information, since it reduces to the mutual information, between the output of the channel and a system purifying its input, in the asymptotic, memoryless scenario.

arXiv:1105.3321•IEEE Transactions on Information Theory•Nov 21, 2012•55 citations

Generalized smooth mutual max-information of quantum channel

Lei Li, Qing‐Wen Wang, Shu‐Qian Shen +1 more

Quantum Information Processing•Jun 7, 2023•0 citations
Underestimating the leakage can compromise physical layer security (PLS), while overestimating it may lead to inefficient system design. Therefore, a reliable leakage estimator is essential. Neural network-based estimators provide a data-driven way to estimate mutual information (M.I.) without requiring full knowledge of the channel or source distributions. In this work, we aim to scale the blocklength of a wiretap code such that the estimator can still feasibly operate. We propose an improved M.I. estimator based on the variational contrastive log-ratio upper bound (vCLUB) framework, tailored for both discrete and continuous variables. By using a mixture of Bernoulli experts parameterized by neural networks, the estimator is able to quantify information leakage in communication systems, which employ complex data processing like universal hash families (UHF). We further propose a method to utilize the proposed estimator to design the UHF for a wiretap code or secret key generation design. Simulation results show that prior methods significantly underestimate the M.I., particularly when using UHF for higher blocklengths (n ≫16). The proposed method can scale the blocklength up to 255, and we conjecture that the design can scale well to even higher blocklengths given adequate training data and model size. Additionally, we contend that our proposed estimator and adaptive hash design framework offer a practical approach for extending PLS considerations for wiretap channels into the finite blocklength regime.
We study entanglement-assisted quantum and classical communication over a single use of a quantum channel, which itself can correspond to a finite number of uses of a channel with arbitrarily correlated noise. We obtain characterizations of the corresponding one-shot capacities by establishing upper and lower bounds on them in terms of the difference of two smoothed entropic quantities. In the case of a memoryless channel, the upper and lower bounds converge to the known single-letter formulas for the corresponding capacities, in the limit of asymptotically many uses of it. Our results imply that the difference of two smoothed entropic quantities characterizing the one-shot entanglement-assisted capacities serves as a one-shot analog of the mutual information, since it reduces to the mutual information, between the output of the channel and a system purifying its input, in the asymptotic, memoryless scenario.