Thursday, 16 January 2025
Hall C (New Orleans Ernest N. Morial Convention Center)
NOAA has been operating environmental satellites in low-Earth orbit for over 50 years. The next generation
of their constellations, the Near-Earth Orbit Network (NEON), is being planned with a shift in design paradigm that
emphasizes quick deployment and scalability, aiming to produce a larger, more resilient constellation of smaller
satellites. This new paradigm requires simultaneous investment in scaling ground systems to support ever-increasing
demand as constellations scale, necessitating intelligent planning to maximize network performance. Traditional
methods of planning ground contacts rely on heuristic approaches or manual scheduling, solutions that do not scale
as the decision space of the problem increases with more satellites and ground stations. The proliferation of artificial
intelligence in computing provides a new alternative to solving this problem. We utilized this AI-enabled approach
when modeling traffic on the NOAA ground station network from several future missions and providers, including the
proposed NEON constellation, to identify and analyze bottlenecks in meeting data latency requirements. To simulate
day-to-day operations, we produced operationally viable ground contact plans targeting optimal data latency while
respecting the pertinent constraints of space and ground assets using CNTIENT-Optimize, our in-house satellite
mission planning software. In this way, we mapped out a trade space between variables such as fleet size, position of
ground assets, data generation and transmission rates, and other operational variables. We conduct a Pareto
analysis to explore the tradeoffs between these key operational variables to enable an understanding of how changes
in one variable affect overall system performance. The performance of the ground services and mission planning
software were evaluated on increasingly complex scenarios, culminating in a 100 satellite fleet, to stress test the
system.
of their constellations, the Near-Earth Orbit Network (NEON), is being planned with a shift in design paradigm that
emphasizes quick deployment and scalability, aiming to produce a larger, more resilient constellation of smaller
satellites. This new paradigm requires simultaneous investment in scaling ground systems to support ever-increasing
demand as constellations scale, necessitating intelligent planning to maximize network performance. Traditional
methods of planning ground contacts rely on heuristic approaches or manual scheduling, solutions that do not scale
as the decision space of the problem increases with more satellites and ground stations. The proliferation of artificial
intelligence in computing provides a new alternative to solving this problem. We utilized this AI-enabled approach
when modeling traffic on the NOAA ground station network from several future missions and providers, including the
proposed NEON constellation, to identify and analyze bottlenecks in meeting data latency requirements. To simulate
day-to-day operations, we produced operationally viable ground contact plans targeting optimal data latency while
respecting the pertinent constraints of space and ground assets using CNTIENT-Optimize, our in-house satellite
mission planning software. In this way, we mapped out a trade space between variables such as fleet size, position of
ground assets, data generation and transmission rates, and other operational variables. We conduct a Pareto
analysis to explore the tradeoffs between these key operational variables to enable an understanding of how changes
in one variable affect overall system performance. The performance of the ground services and mission planning
software were evaluated on increasingly complex scenarios, culminating in a 100 satellite fleet, to stress test the
system.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner