Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Scalable Scheduling Policies for Quantum Satellite Networks

Albert Williams, Nitish K. Panigrahy, Andrew McGregor, Don Towsley

TL;DR

This paper addresses scheduling entanglement distribution in quantum satellite networks under satellite and ground-station resource constraints, formalizing the Quantum Satellite Scheduling Problem ($QSSP$). It proves $QSSP$ is $NP$-hard via a reduction from $3$-Dimensional Matching and then proposes four heuristics—$RANDOM$, $LOCAL\_GREEDY$, $GLOBAL\_GREEDY$, and $GREEDY\_BACKOFF$—along with a simulation platform to evaluate them on Starlink-like mega-constellations using a dual-downlink, no-onboard-memory architecture. The study shows that $GREEDY\_BACKOFF$ and $GLOBAL\_GREEDY$ consistently deliver higher entanglement distribution rates and fidelities, with robustness to ground-station receiver limits and time-of-day effects. This work provides a scalable framework and software tools for evaluating large-scale quantum-satellite networks, informing practical deployment and future enhancements such as memory-enabled satellites and inter-satellite links.

Abstract

As Low Earth Orbit (LEO) satellite mega constellations continue to be deployed for satellite internet and recent successful experiments in satellite-based quantum entanglement distribution emerge, a natural question arises: How should we coordinate transmissions and design scalable scheduling policies for a quantum satellite internet? In this work, we consider the problem of transmission scheduling in quantum satellite networks subject to resource constraints at the satellites and ground stations. We show that the most general problem of assigning satellites to ground station pairs for entanglement distribution is NP-hard. We then propose four heuristic algorithms and evaluate their performance for Starlink mega constellation under various amount of resources and placements of the ground stations. We find that the maximum number of receivers necessary per ground station grows very slowly with the total number of deployed ground stations. Our proposed algorithms, leveraging optimal weighted b-matching and the global greedy heuristic, outperform others in entanglement distribution rate, entanglement fidelity, and handover cost metrics. While we develop these scheduling algorithms, we have also designed a software system to simulate, visualize, and evaluate satellite mega-constellations for entanglement distribution.

Scalable Scheduling Policies for Quantum Satellite Networks

TL;DR

This paper addresses scheduling entanglement distribution in quantum satellite networks under satellite and ground-station resource constraints, formalizing the Quantum Satellite Scheduling Problem (). It proves is -hard via a reduction from -Dimensional Matching and then proposes four heuristics—, , , and —along with a simulation platform to evaluate them on Starlink-like mega-constellations using a dual-downlink, no-onboard-memory architecture. The study shows that and consistently deliver higher entanglement distribution rates and fidelities, with robustness to ground-station receiver limits and time-of-day effects. This work provides a scalable framework and software tools for evaluating large-scale quantum-satellite networks, informing practical deployment and future enhancements such as memory-enabled satellites and inter-satellite links.

Abstract

As Low Earth Orbit (LEO) satellite mega constellations continue to be deployed for satellite internet and recent successful experiments in satellite-based quantum entanglement distribution emerge, a natural question arises: How should we coordinate transmissions and design scalable scheduling policies for a quantum satellite internet? In this work, we consider the problem of transmission scheduling in quantum satellite networks subject to resource constraints at the satellites and ground stations. We show that the most general problem of assigning satellites to ground station pairs for entanglement distribution is NP-hard. We then propose four heuristic algorithms and evaluate their performance for Starlink mega constellation under various amount of resources and placements of the ground stations. We find that the maximum number of receivers necessary per ground station grows very slowly with the total number of deployed ground stations. Our proposed algorithms, leveraging optimal weighted b-matching and the global greedy heuristic, outperform others in entanglement distribution rate, entanglement fidelity, and handover cost metrics. While we develop these scheduling algorithms, we have also designed a software system to simulate, visualize, and evaluate satellite mega-constellations for entanglement distribution.
Paper Structure (23 sections, 1 theorem, 8 equations, 4 figures, 2 tables, 2 algorithms)

This paper contains 23 sections, 1 theorem, 8 equations, 4 figures, 2 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Preliminaries
  3. Satellites and Ground Stations
  4. Dual Downlink Architecture
  5. Entanglement Generation Source
  6. Loss and Noise
  7. Scheduling in a time snapshot
  8. Quantum Satellite Scheduling Problem
  9. NP-hardness of QSSP
  10. Special Cases
  11. Satellites outnumbering ground stations
  12. No resource constraints at ground stations
  13. Heuristic Algorithms
  14. Performance Evaluation
  15. The Starlink constellation
  16. ...and 8 more sections

Key Result

Theorem 3.1

The Quantum Satellite Scheduling Problem is NP-hard.

Figures (4)

  • Figure 1: The Dual Downlink architecture for entanglement distribution.
  • Figure 2: Satellites and ground station distributions, $|S| = 3967, |G| = 100.$ (a) Satellites in the Starlink LEO mega-constellation. (b) Ground stations are placed uniformly at random on the land. (c) Ground stations are chosen based on highest population density.
  • Figure 3: (a) Satellite altitudes and inclination angles in the Starlink LEO mega-constellation. (b) Maximum number of satellites visible per ground station pair over time. (c) Maximum number of ground station pairs visible per satellite over time. (d) Entanglement distribution rates achieved by different heuristics. (e) Average number of connections per ground station for different heuristics. (f)-(g) Effect of number of receivers on performance. (h)-(i) Longevity of the assigned connections. (j)-(k) Average fidelity vs number of receivers across day time and night time. (l) Effect of different time of the year on entanglement fidelity.
  • Figure 4: Ground stations in Asia connected with LOCAL_GREEDY, and GREEDY_BACKOFF. Red links are the connections assigned only by GREEDY_BACKOFF, blue links are assigned only by LOCAL_GREEDY, and black links are assigned by both.

Theorems & Definitions (2)

  • Theorem 3.1
  • Remark 1