How ghinwaismail/multimodal-traffic-digital-twins works
This is a notebook-first research artifact for trace-driven burst traffic generation for 5G digital twins. The documented per-application workflow preprocesses packet traces, fits a clustering/Markov burst model, generates synthetic bursts, and exports results; a separate notebook consolidates paper figures and tables. The tree includes several application notebooks and comparison outputs. Source excerpts were not provided, so relationships below rely on the README; undocumented implementation details are omitted.
Main entry points: comparison_figures.ipynb Paper reporting · notebooks Preprocess traces · comparison_figures Comparison results · comparison_statistics_downlink.csv Downlink statistics
Components 15
Traffic pipeline
- Preprocess traces
notebooks - Fit burst model
notebooks - Generate bursts
notebooks - Export results
notebooks - Aparat pipeline
notebooks/Traffic_generator_aparat.ipynb - Filimo pipeline
notebooks/Traffic_generator_filimo.ipynb - iGap pipeline
notebooks/Traffic_generator_igap.ipynb - Telegram pipeline
notebooks/Traffic_generator_telegram.ipynb - YouTube pipeline
notebooks/Traffic_generator_youtube.ipynb - Multi-model pipeline
notebooks/Traffic_generation_multi_model.ipynb
Paper reporting
- Paper reporting
notebooks/comparison_figures.ipynb - Comparison results
comparison_figures - Downlink statistics
comparison_figures/comparison_statistics_downlink.csv - Comparison figures
comparison_figures
Other components
- Packet traces
Connections 7
- Packet traces → Preprocess traces: provides input
README.md - Preprocess traces → Fit burst model: precedes fitting
README.md - Fit burst model → Generate bursts: precedes generation
README.md - Generate bursts → Export results: precedes export
README.md - Paper reporting → Comparison results: consolidates results
README.md - Paper reporting → Downlink statistics: exports tables
README.md - Paper reporting → Comparison figures: exports figures
README.md