quant-phMar 3, 2026

Optimizing Orbital Parameters of Satellites for a Global Quantum Network

Authors: Athul Ashok, Owen DePoint, Jackson MacDonald, Albert Williams, Don Towsley

Organizations: University of Massachusetts, Amherst, MA, USA

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Due to fundamental limitations on terrestrial quantum links, satellites have received considerable attention for their potential as entanglement generation sources in a global quantum internet. In this work, we focus on the problem of designing a constellation of satellites for such a quantum network. We find satellite inclination angles and satellite cluster allocations to achieve maximal entanglement generation rates to fixed sets of globally distributed ground stations. Exploring two black-box optimization frameworks: a Bayesian Optimization (BO) approach and a Genetic Algorithm (GA) approach, we find comparable results, indicating their effectiveness for this optimization task. While GA and BO often perform remarkably similar, BO often converges more efficiently, while later growth noted in GAs is indicative of less susceptibility towards local maxima. In either case, they offer substantial improvements over naive approaches that maximize coverage with respect to ground station placement.

Figures & tables

Fig. 1 : Example of a constellation with the hyperparameter D=3D=3 and the following parameters: θ1=25∘,θ2=75∘,θ3=150∘,x1=2,x2=3,x3=5\theta_{1}=25^{\circ},\theta_{2}=75^{\circ},\theta_{3}=150^{\circ},x_{1}=2,x_{2}=3,x_{3}=5
Fig. 3 : An example of a Gaussian Process
Fig. 4 : Map visualizations of the optimal satellite configurations yielded by GA (red) and BO (purple) for 2 orbits over population clusters.
Fig. 5 : Performance comparison of optimization methods and baselines across different orbit counts and ground station distributions. (a)–(b): Average entanglement distribution rate (EPR pairs/second) across 1, 2, 3, and 5 orbits using population-based (a) and random (b) land-based ground station placements. The single orbit case benchmarks against a brute-force baseline, while multi-orbit results are compared against equispaced inclination angles. (c)–(d): Optimization progress over simulation calls for population-based (c) and random (d) ground stations, showing the maximum rate by the time of a call. (e): Comparative performance across methods (BO, GA, equispaced) under different ground station distributions for 5 orbits. (f): Simulation call at which each optimization method first achieved its final best solution, compared across orbit counts (2, 3, and 5) and ground station distribution types (random and population-based).

Explore similar work

CardsList

  1. SatQNet: Satellite-assisted Quantum Network Entanglement Routing Using Directed Line Graph Neural Networks

    Apr 10, 2026Tobias Meuser, Jannis Weil, Aninda Lahiri +1Graph Neural Networks

    Communication Networks Lab, Technical University of Darmstadt · Institute of Communication Technology, Leibniz University Hannover · Quantum Information Group, IMDEA Networks

  2. Free-Placement Optimization of Ground Station Locations for Low-Earth Orbit Satellites

    Jun 10, 2026Grace Ra Kim, Duncan Eddy, Vedant Srinivas +1Combinatorial OptimizationWireless Communications

    Stanford University, Stanford, CA, 94305 · Stanford, CA · Stanford University, Stanford, CA

  3. Differentiable Satellite Constellation Configuration via Relaxed Coverage and Revisit Objectives

    Apr 21, 2026Shreeyam Kacker, Kerri CahoyDifferentiable ProgrammingDifferentiable Optimization

    Department of Aeronautics and Astronautics MIT