Supporting data for SPARC: Resource-first container orchestration for caching in dynamic LEO satellite networks

Description

This dataset accompanies “SPARC: Resource-first container orchestration for caching in dynamic LEO satellite networks.” It contains recorded simulation result tables, experiment configurations, audit summaries, metric definitions, and scripts for numerical verification and reproduction of Figures 5–8. The records cover the single-type capacity comparison and resource-plane ablations, a separate mobility study, multi-type locality-weight sweeps, and controller-scaling measurements. The figure-provenance map identifies the source tables, while the data dictionary distinguishes intermediate allocation targets, accepted plans, and physical-storage measurements. All retained numerical records are unchanged; no experiments were rerun during manuscript revision. This is a numerical supplement rather than a complete executable reproduction package: the full simulator, deployment environment, Docker images, raw request traces, and raw execution logs are not included.

Steps to reproduce

1、Extract the dataset and open a terminal in the directory containing verify_results.py. 2、Use Python 3 with Matplotlib and NumPy. The plotting environment is recorded in plotting_environment.json. 3、Run python verify_results.py to check exported quota bounds, request-ground accounting, conditional-latency mixtures, protected-class plan/storage agreement, and the policy-weighted objective. 4、Run python tools/plot_final_validation_manuscript.py --output-dir figures to regenerate Figure_5.pdf through Figure_8.pdf. 5、Consult README.md, DATA_DICTIONARY.md, and figure_provenance.json for definitions and experiment boundaries. These commands read recorded results; they do not rerun simulations.

Institutions

• Fudan University

  Shanghai, Shanghai

Categories

Licence