On the Achievable Rate of Satellite Quantum Communication Channel using Deep Autoencoder Gaussian Mixture Model

TL;DR

This paper compares GMM and DAGMM models for estimating satellite quantum channel capacity, demonstrating that DAGMM offers tighter bounds and better noise modeling, advancing quantum satellite communication analysis.

Contribution

It introduces the Deep Autoencoder Gaussian Mixture Model (DAGMM) for improved capacity estimation in quantum satellite channels, surpassing traditional GMM methods.

Findings

DAGMM provides tighter capacity bounds than GMM.

DAGMM captures non-linear noise variations more effectively.

The study introduces the Deep Cluster Gaussian Mixture Model (DCGMM) for high-dimensional quantum data.

Abstract

We present a comparative study of the Gaussian mixture model (GMM) and the Deep Autoencoder Gaussian Mixture Model (DAGMM) for estimating satellite quantum channel capacity, considering hybrid quantum noise (HQN) and transmission constraints. While GMM is simple and interpretable, DAGMM better captures non-linear variations and noise distributions. Simulations show that DAGMM provides tighter capacity bounds and improved clustering. This introduces the Deep Cluster Gaussian Mixture Model (DCGMM) for high-dimensional quantum data analysis in quantum satellite communication.