Transport on Intermittent Networks
The Timetable Is the Network
The Argument
- Problem:
- Delay-tolerant networks have no rigorous reliability certification method.
- Core result:
- Deliverability decomposes exactly as DR = ST × η. (Arithmetic identity — true by construction.)
- Classification:
- The order parameter γ separates temporal contact graphs into two classes (cluster vs. trap) with zero sign overlap across 155,000+ configurations. (Empirical regularity — strong evidence, no proof.)
- Applied consequence:
- The current cislunar architecture fails certification in the data-supported parameter range. (Inference from simulation results.)
- What would kill it:
- A temporal contact graph where γ > 0 but delivery ratio decreases with added path diversity. None found in 290,000+ configurations. (Open falsification target.)
What This Means in Practice
Of every 100 tonnes of cryogenic propellant dispatched from Earth toward Jupiter via relay chain, fewer than 3 arrive. This is not an insulation problem or a budget problem. It is a consequence of orbital mechanics and hydrogen thermodynamics that no relay architecture with passive cryogenic storage can overcome.
Interactive Tools
Sparse Law Calculator
Predict delivery ratios from network parameters using the three-factor sparse law. 17,280 configurations embedded.
Solar System Orrery
3D explorer with Keplerian orbital mechanics, DTN contact links, and constellation overlays.
Papers
The foundational paper. Factorization, classification, 155,000+ configurations.
- From classification to mechanism in temporal contact networks — why networks trap: Var[ln p] mechanism, four-route taxonomy, layered diagnostic hierarchy
- Percolation-optimal relay architectures for cislunar and interplanetary networks — 89,178 configurations, Braess paradox, design rules for Artemis/Gateway and Mars Sample Return
- Jβ saturation invariance and TASEP phase structure in DTN relay networks — load-dependent phase transitions, universal current ceiling, golden-mean falsification
- Policy brief: Jupiter cryogenic propellant — two-page summary of implications for relay architecture planning
Code & Data
- github.com/toxic2040/TIN — source code, simulation engine, 298 experiment runners
- License: AGPL-3.0-or-later
About
J. Councilman is an independent researcher applying percolation theory to scheduled transport networks. The core method: decompose aggregate metrics into their upstream causes before reaching for complex explanations. 290,000+ configurations tested, all code open-source.