Swarm Coordination Architecture Simulator

Monte Carlo simulation comparing coordination topologies for large-scale Dyson swarms. Analyze communication overhead, latency, and availability across centralized, hierarchical, and mesh network architectures at scales up to 1 million nodes.

Swarm Parameters

1k 500k 1M
Coordination Topology
50 125 200
1h 84h 168h (1 week)
0.1 5 10
20 35 50
0 (quiescent) 0.10 (reconfig) 1.0 (stress)
1% 5% 10%

Comparison Mode

Compares all three topologies (centralized, hierarchical, mesh) with the same base parameters to identify the optimal architecture for your swarm size.

Research Questions

  • rq-1-24: Architecture at scale
  • rq-1-39: Coordinator duty cycle
  • rq-2-17: Fleet coordination constraints

Topology Comparison

Run a comparison simulation to see topology chart

TDMA Feasibility (Paper Model)

INFEASIBLE at 35 kbps

gamma = 0.732 | eta_total = 30.2% | margin = -328 ms

Test A: Byte Budget

PASS

Baseline telemetry20.5%
Architecture5.6%
Command (d=0.1)4.1%
Total30.2%

Test B: TDMA Timing

FAIL

Ingress7917 ms
ARQ (P95)2239 ms
Egress172 ms
Margin-328 ms
Info rate: 20.3 kbps
Min PHY: 35.0 kbps
alpha_RX: 0.792

Simulation Results

Configure the swarm parameters and run the simulation to see results.

Simulation Methodology

This discrete event simulation models coordination patterns for Dyson swarm elements, comparing three network topologies: centralized (single coordinator), hierarchical (cluster-based with regional coordinators), and mesh (peer-to-peer gossip protocol).

  • Message passing simulates ephemeris updates, handoffs, and collision warnings
  • Coordinator duty cycling models power constraints and rotation schedules
  • Node failures follow exponential distribution based on annual failure rate
  • Light-time delays estimated for swarm orbital positions

Key trade-off: Centralized has lowest latency but single point of failure; hierarchical balances scalability with structure; mesh provides best fault tolerance but higher latency.

Project Dyson — A volunteer-led nonprofit. All plans and research are publicly available.