Optimization of Satellite Constellation Design and Network Topology
Name
capper-jcapper-ms-aeroastro-2026-thesis.pdf
Size
8.26 MB
Format
Adobe PDF
Checksum (MD5)
5c363c45226daf86d910684aa46fe012
Author(s)
Capper, Jack
Advisor(s)
Modiano, Eytan
Date Issued
February 2026
Publisher
Massachusetts Institute of Technology
Abstract
The rapid growth satellite services demand, along with the accompanying proliferation of satellite constellations, has unveiled the need for optimization methods to improve satellite network topologies and constellation designs to provide more efficient and cost effective constellations while still meeting the ever increasing demand. This thesis proposes and evaluates novel optimization frameworks for satellite network topologies, as well as constellation design. The proposed methodologies leverage Genetic Algorithms to achieve flexible, realizable systems tailored to specific mission requirements. The first of the two methodologies implements both static and dynamic satellite network topology optimization. The static framework minimizes Average Shortest Path Length at a single snapshot in time across the satellite network, with the option to include ground stations, while the dynamic framework optimizes jointly over multiple time steps and incorporates penalties for link changes to balance link churn with performance improvements over time. Additionally, a joint constellation design optimization framework is proposed, employing a pseudo variable-length chromosome within a multi-objective Genetic Algorithm to optimize satellite constellations across multiple orbital regimes. This approach enables exploration of diverse design spaces with no restriction on constellation designs or the number of orbital shells, while considering realistic constraints and objectives such as coverage, latency, launch cost, and resilience. Extensive simulation demonstrates the performance of both the proposed methodologies. Results reveal key trends, including a relationship between the optimal vertex-symmetric topology and the constellation’s size and shape for Walker Delta constellations, the benefits of asymmetric topologies in reducing ASPL, and the trade-offs between topology stability and dynamic reconfiguration. Furthermore, optimized constellation designs are shown to outperform traditional designs through a comparison with a set of baseline constellations. Several objective trade-offs present in the resultant pareto fronts are also explored in depth. The proposed optimization methodologies in this thesis enable the development of more efficient, higher performing, and cost-effective satellite constellations. This area of research is also ripe for future work, with potential to focus on incorporating traffic-aware objectives, link-level impairments, constellation augmentation or reconfiguration, and additional optimization metrics to further enhance the operational applicability of the optimization results.
MIT Department
Massachusetts Institute of Technology. Department of Aeronautics and Astronautics
Terms of Use
In Copyright - Educational Use Permitted
Copyright retained by author(s)
Persistent DSpace Link
