English | Español

---

This project provides a reproducible corpus for DTN/OppNets research in The ONE simulator, plus a full analysis pipeline and documentation stack for paper writing.

Current paper-ready status:

• 60 scenarios across 7 families
• Diversity analysis in full-46 and core-23 descriptor spaces
• Freeze declared as final optimized baseline (publishable, not optimal final corpus)

---

Metric	Value
full-46 max |r|	0.9377
full-46 pairs |r| >= 0.7	46/1770 (2.6%)
full-46 min cosine	0.0620
full-46 silhouette (Ward k=7)	0.2929
core-23 max |r|	0.9829
core-23 pairs |r| >= 0.7	58/1770 (3.3%)
core-23 silhouette (Ward k=7)	0.2681
feature-feature core	mm_WDM <-> mm_Bus = 0.9393

Official frozen numbers are maintained in Final-frozen-results.

---

• Build a scenario set that is broad enough to benchmark routing protocols without linear redundancy collapse.
• Keep a transparent methodology (features, NaN/normalization, correlation, clustering, ablation).
• Preserve traceability from scenario settings to analysis artifacts and conclusions.

---

• Research-goals
• Thesis-phases
• Contributions
• Repository-structure

• Corpus-overview
• Scenario-families
• Corpus-versioning
• Dropped-scenarios-evolution
• Scenario-catalog

• Scenario-representation
• Core-vs-extended-features
• Feature-selection-rationale
• NaN-and-normalization-policy
• Diversity-analysis-methodology
• Marginal-test
• Ablation-methodology
• Methodological-limitations

• Results-overview
• Final-frozen-results
• Optimization-history
• Feature-feature-analysis
• Clustering-analysis
• Output-space-analysis
• Figures

• Quickstart
• Running-analysis-pipeline
• Generating-figures
• Using-corpus-in-the-one
• Data-and-artifacts