QoS-aware user grouping and beam placement for resource constrained leo satellite downlinks

Date

2026-6-26

Author

Armutcu, İsmail Hakkı

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

39
views

0
downloads

Cite This

This thesis addresses QoS-aware user clustering and beam placement for resource-constrained low Earth orbit (LEO) satellite downlinks. The problem is modeled as a static regional snapshot in which users with heterogeneous traffic demands and quality-of-service (QoS) classes are assigned to visible satellites and grouped into steerable spot beams. Feasible plans must satisfy visibility, coverage, link-rate, per-beam load, beam-count, and aggregate payload constraints. The thesis proposes QoS-Aware Split-Refine, an interpretable heuristic that combines load-aware satellite association, split-to-feasible beam construction, QoS-sensitive beam center placement, edge-risk refinement, load balancing, and payload repair. QoS-Aware Split-Refine is evaluated against graph-based and k-means-style baselines, design variants, ablations, sensitivity studies, and a small-instance mixed-integer linear programming (MILP) diagnostic benchmark. Results show that QoS-Aware Split-Refine keeps high-priority users farther from weak beam-edge regions, achieves the lowest average peak utilization among the compared methods, and preserves payload feasibility. Ablations show that QoS refinement drives placement gains for high-priority users, while load balancing improves utilization. MILP and runtime analyses indicate that a grid-discretized MILP benchmark is useful for small diagnostic cases, whereas QoS-Aware Split-Refine remains practical at regional scale. The thesis positions QoS-aware beam clustering as a scalable intermediate layer between geometric clustering and full dynamic resource allocation.

Subject Keywords

LEO satellite downlinks, QoS-aware user clustering, Spot beam placement, Satellite resource management, Payload constraints

URI

https://hdl.handle.net/11511/119788

Collections

Graduate School of Natural and Applied Sciences, Thesis