Reproducibility Package for Latency-Aware Intelligent Offloading in Integrated Space–Terrestrial Vehicular Edge Networks

Published: 13 October 2026| Version 1 | DOI: 10.17632/9shn8mmnz7.1
Contributors:
,
,
,

Description

This dataset contains the reproducibility package associated with the manuscript “Latency-Aware Intelligent Offloading in Integrated Space–Terrestrial Vehicular Edge Networks.” The package includes simulation and statistical-analysis notebooks for dynamic spectrum access, task-size service delay, task/load redistribution, satellite OTFS-gain sensitivity, and offloading-method comparison, together with dataset-generation, DNN training and evaluation, and CPU timing code for the offloading and task-division models. Generated CSV results, trained-model artifacts, timing bundles, and figures supporting the results reported in the manuscript and supplementary material are included. The accompanying README documents the package structure, software environment, execution conventions, and mapping between the individual packages and manuscript results. The included datasets and simulation outputs are synthetically generated from the system models and parameter settings described in the manuscript.

Files

Steps to reproduce

See README.md for the complete package structure and execution guidance. Install the Python dependencies listed in requirements.txt using Python 3.11.5. Each experiment or learning-model package is self-contained and should be run from its own directory because relative paths are used for generated or stored files. For simulation packages, run the corresponding simulate_*.ipynb notebook first and then the matching analyze_*.ipynb notebook where provided. The DNN packages include separate dataset-generation, model-training/evaluation, and CPU-timing components; the README specifies the corresponding execution order and the purpose of the included frozen timing bundles. Generated result files are included so the reported outputs can also be inspected without rerunning the computational experiments.

Institutions

Categories

Computer Science, Computer Communications

Licence