Documentation map
Where to find things in docs/. Repo-level orientation is one directory up:
README.md (front door), CONTEXT.md
(project orientation + glossary), AGENTS.md (build/verify,
golden rules, where-to-look), CONTRIBUTING.md.
Understand the protocol
| Page |
What it is |
| ant-colony-routing.md |
Concepts primer: ant foraging → ACO → AntNet → AntHocNet. Start here for the idea. |
| ant-types.md |
Reference for the five ant types: comparison table, lifecycle diagrams (setup, maintenance, repair), and how to observe them at runtime. |
| architecture.md |
The core/adapter split, ports, and the decision flow. |
| roadmap.md |
Release ladder v1.3.0→v3.0.0, the epic dependency graph, per-release exit criteria, and the deliberate non-goals. Live status lives on #298. |
| software-layers.md |
Three diagrams: the software stack, the ant mechanisms + the switches that gate them, and what is live/inert/planned per network regime. |
| network-regimes.md |
Why MANET and satellite/ISL are different routing problems — read before transferring an intuition between them. §6 tables which AntHocNet mechanism is live/inert in each regime. |
Build, port, extend
| Page |
What it is |
| porting-notes.md |
Bugs fixed in extraction, NS-2 patch anchors, version caveats. |
| ns2-support.md |
The deprecated NS-2 target in one place — frozen at v1.2.0, still ships through v1.x, removed at v2.0.0 (#307); install, images, rationale. |
| wire-format.md |
Canonical on-wire ant layout and the kWireVersion rules (golden rule 4). |
| configuration.md |
Every tunable, its default's provenance, the ns-3/NS-2 surface for it, and the calibration loop. |
| cross-validation.md |
NS-2 vs NS-3 behaviour re-validation (not bit-for-bit parity). |
Measure & reproduce
| Page |
What it is |
| benchmarks.md |
Results index: per-merge taxonomy table + links to every scenario/sweep page. |
| benchmarks/metrics.md |
What PDR, delay99, NRL etc. mean, and their caveats. |
| benchmarks/methodology.md |
Reproduce commands (local and CI dispatch), build profiles, validation anchors, determinism gate, and what each baseline arm is for. |
| benchmarks/modern-baseline-survey.md |
Dated survey of ns-3 support for Babel / BATMAN-adv / OLSRv2, and the decision it drove. |
| benchmarks/README.md |
How the figures/tables are generated and regenerated. |
| benchmarks/satellite/isl-grid.md |
The satellite/ISL suite: harness, analytic anchors, dispatch. |
| benchmarks/grid.md |
The mobility × channel grid and its scoped ranking-stability statement. |
| benchmarks/reinjection.md |
The #46 MAC-failure detector arm: the 20-seed detector A/B, the duplicate-re-injection measurement, and the cap frontier. |
Research provenance & fidelity
| Page |
What it is |
| fidelity.md |
What this implementation reproduces from the 2004 paper, and every deliberate deviation — the claims page. |
| publications/ |
Source-of-truth digests of the 2004 PPSN paper and the 2007 Ducatelle thesis (no vendored PDFs). |
| satellite-routing-prior-art.md |
ACO-on-constellations survey behind the satellite track (#192). |
Decisions & process
| Page |
What it is |
| adr/ |
Architecture Decision Records 0001–0020, indexed with one-line summaries. |
| handoffs/ |
Dated cross-session investigation handoffs (see ADR-0013 for the issue-first discipline). |