EX06 — Steered Multibeam Service¶

Tags: multibeam · dynamic targeting · regional demand Input: steered-multibeam.yaml · Runtime tier: standard · Recommended views: 2D, 3D, Analysis

Purpose¶

Explore how an eight-beam payload layout and periodic demand-aware steering turn regional demand into selected beam targets and delivered service.

Setup¶

Twelve satellites serve western and eastern regions through Madrid and Algiers over four hours. Each spacecraft carries an elongated eight-beam layout.

astraeus run astraeus-resources/examples/user/steered-multibeam.yaml

RunSpec Focus¶

The targeting policy refreshes every ten minutes, ranks cells using service population demand, and preserves a ten-minute minimum dwell.

transport:
  beams:
    variant: steered
    count_per_satellite: 8
    targeting:
      mode: dynamic_demand_weighted
      demand_source: service_population
      target_count: 8
      min_dwell_s: 600
      refresh:
        interval_s: 600

Results¶

The reference run completed successfully in 1.09 seconds. It generated 200 target rows across eight unique target cells: 147 were applied and 53 rejected by the physical targeting evaluation. The scheduler delivered 88.8 MB of 352.8 MB admitted traffic (25.17%), with a 41.43 ms p95 served-path latency.

Targeting or service outcome	Result
Unique demand cells targeted	8
Requested target rows	200
Applied target rows	147
Rejected target rows	53
Delivered traffic	88.8 MB
Byte delivery rate	25.17%

Artifacts To Inspect¶

Artifact	What it shows
`transport/beam_targeting_summary.json`	Target counts, application outcomes, and optimizer diagnostics
`transport/beam_target_plan.parquet`	Time-resolved target selected for every steerable payload beam
`transport/beam_target_link_outcome_rollup.parquet`	Physical link closure for the requested targets
`mission/service_population_beam_target_assignments.parquet`	Connection between resolved demand and beam targets

Interpretation¶

Demand-aware targeting chooses where payload opportunity should be concentrated; it does not guarantee that every requested pointing closes. The applied and rejected target rows connect the steering objective to physical feasibility, while the service ledger shows the resulting delivery.