Constellation Design Optimization: Coverage, Revisit, and Cost Trade-Offs

When designing a satellite constellation, the foundational challenge is a multi-dimensional optimization problem: achieve required coverage and revisit performance while minimizing total mission cost. Every design decision (orbital altitude, inclination, number of planes, satellites per plane) has cascading implications for cost, latency, launch mass, radiation environment, and link budget. This guide introduces the core analytical framework.

Key Performance Parameters

Before optimizing, you need to define what you're optimizing for. Constellation performance is typically characterized by:

• Coverage: The fraction of Earth's surface (or a target region) visible to at least one satellite at any given time. Continuous global coverage requires many satellites; regional coverage with gaps can be achieved with far fewer
• Revisit time: For non-continuous-coverage constellations, the time between successive observations of a fixed ground point. Daily revisit requires a different architecture than hourly revisit
• Latency: For communications constellations, the delay between ground terminal and satellite hop. Lower-altitude orbits reduce latency but increase the number of satellites required for global coverage
• Elevation angle: The minimum angle above the horizon for a satellite to be usable. Higher minimum elevation (e.g., 30° for reliable communications) increases the satellite count needed for coverage
• Capacity: For communications, the aggregate throughput per unit area or per user, a function of frequency, bandwidth, and satellite EIRP

Altitude Selection

Altitude is the primary architectural decision. The key trade space:

• Low Earth Orbit (LEO, 200–2,000 km): Low latency (20–40 ms for communications), low free-space path loss, but many satellites required for global coverage and short orbital lifetimes below ~600 km. Avoids the inner Van Allen belt. The dominant choice for broadband (Starlink, Kuiper) and Earth observation constellations
• Medium Earth Orbit (MEO, 2,000–35,000 km): Each satellite sees a larger fraction of Earth, so fewer are needed. GPS, Galileo, and GNSS systems operate at ~20,200 km for this reason. Higher radiation environment (Van Allen belts); components must be radiation-hardened
• Geosynchronous Orbit (GEO, 35,786 km): Three satellites can cover nearly all of Earth (minus polar regions). Very high latency (~600 ms round-trip), large antenna required on the ground side, and high launch cost. Optimal for broadcast, weather, and early communication missions

Walker Constellations

The standard analytical framework for symmetric constellations is the Walker delta pattern, specified by the notation T/P/F where:

• T = total number of satellites
• P = number of orbital planes
• F = relative phasing between planes (0 ≤ F < P)

All planes have the same inclination, altitude, and eccentricity (circular). Satellites in each plane are equally spaced. The Walker framework allows systematic coverage analysis: for a given altitude and inclination, you can compute the minimum T/P/F that achieves continuous global coverage at a specified minimum elevation angle.

For example, continuous global coverage at 45° inclination from 1,200 km altitude with a 10° minimum elevation angle requires on the order of 40–60 satellites in a Walker delta configuration, depending on exact inclination and phasing choices.

Inclination Trade-Offs

• Polar (90°) and near-polar (~98° sun-synchronous): Provides coverage over the entire Earth including poles. Each satellite covers all latitudes. Required for global Earth observation missions. Higher inclination raises launch cost from equatorial launch sites (plane change penalty)
• High inclination (50–70°): Good mid-latitude coverage. Starlink uses ~53° inclination for its primary shell, providing strong coverage for the majority of the world's population at mid-latitudes while reducing the satellite count needed versus polar
• Low inclination (28–45°): Concentrates coverage at lower latitudes (equatorial and subtropical regions). Most efficient per satellite for a specific target latitude band. Less useful for global service

Cost Drivers

Constellation cost is dominated by:

• Satellite unit cost × count: Scaling to hundreds or thousands of satellites requires mass-production economics. Bus and payload standardization, supply chain management, and production rate are critical
• Launch cost: Rideshare and dedicated small launch options have dramatically reduced per-satellite launch cost, but the total launch manifest for a large constellation is still substantial. Targeting altitudes accessible by Falcon 9 rideshare (e.g., 550 km) reduces this cost
• Replacement rate: Satellites have finite lifetimes (3–7 years typical for LEO smallsats). A 200-satellite constellation at 5-year average life requires 40 replacement satellites per year, a significant ongoing cost
• Ground segment: Gateway stations, control centers, and user terminals; often comparable in total cost to the space segment over the mission lifetime

Use the SpaceNexus Constellation Designer to model coverage, revisit time, and satellite count for custom Walker configurations. Combine with the Launch Cost Calculator for rough program cost estimates.