home
spacetime#
A Simulation of Heterogeneous Networked Lunar Clocks.
Abstract#
This project aims to use agent-based modeling and simulations to explore how network topology and time synchronization evolve in a growing lunar ecosystem. Modeling how the approach may be used in practice and at scale informs decisions about how to architect a robust cislunar PNT system that scales with the population.
Purpose#
- The code in this repository is intended for analytical use for a specific problem. It is not meant to be a general-purpose library.
- This project and its code are evolving as my personal understanding of modeling and coding techniques on the way to discovering answers to questions about the nature of a possible decentralized PNT service for the Moon.
Problem statement#
The effectiveness of Decentralized PNT versus traditional Lunar GNSS is only meaningful if we can evaluate both paradigms (independently or concurrently) at different scales in population.
Goals | Apples-to-apples comparisons in performance of a decentralized, centralized, peer-to-peer, transmitter/receiver, and combinations therein based on quantitative metrics. |
---|---|
Non-goals | This study does not intend to discern optimal constellation architectures or compositions. It also does not claim to produce true-to-life simulations of system performance. |
The model shall…#
- depend only on open source software.
- simulate populations of heterogeneous agents.
- simulate interactions between tens of agents and thousands of agents.
- simulate events over a time span of days and a span of years.
- allow an agent’s perception of the environment (epoch, motion) to differ from the “true” environment parameters.
- simulate spacetime (motion, time with relativity, light propagation) on a cislunar scale.
- simulate things that inhibit interactions like distance (attenuation, inverse square law) or obstructions (line of sight blocked by a planetary body).
- allow rulesets for permitted interactions between agents (faction dynamics, transmit/receive limitations, data transfer vs passive sensing, limited communication spectral bands).
The model should…#
- feed visualizations of populations by type, position, and/or interactions.
- execute in less than an hour on my laptop.
- permit rapid execution loops or batched runs with varied model parameters.
- be easy to replicate by others and on other machines.