Albert Vasso, Air Force Institute of Technology (AFIT); Richard Cobb, Air Force Institute of Technology (AFIT); John Colombi, Air Force Institute of Technology (AFIT); Bryan Little, Air Force Institute of Technology (AFIT); David Meyer, Air Force Institute of Technology (AFIT)
Keywords: SDA, SSA, non-traditional, modeling and simulation, system architecting
Abstract:
Space Domain Awareness (SDA) requirements are increasing while the United States Government (USG), the
worlds largest provider of SDA and Space Traffic Management (STM) services, struggles to update legacy hardware
and address new threats. Recent policy asserts the need to utilize non-traditional sources to improve these mission
areas, but no study has approached the problem as the optimization of a new, augmented USG/non-traditional network
(AN). This multi-disciplinary study explores the problem by employing system architecting, physics-based modeling,
optimization, and data analysis to resolve a hypothetical System Program Office (SPO) contracting decision. The
SPO is charged with designing the AN, which is composed of the three Ground-Based Electro-Optical Deep Space
Surveillance (GEODSS) systems, one contributing civil telescope, one contributing large allied scientific telescope,
and some number of fully-taskable commercial small telescopes. The SPO must decide which of the 56 worldwide
commercial sensors to purchase from three companies given a total cost constraint of $25M. Literature review and
market research determined representative non-traditional capabilities while system architecting identified coverage,
average capacity, and average latency to be amongst the most important measures of AN performance. A large-scale
tradestudy exploring the 1016 possible AN architectures was conducted by modeling architectures and 954
Geosynchronous Earth Orbit (GEO) Resident Space Objects (RSOs) in Systems Tool Kit (STK) and custom Python
scripts, then simulating architecture performance over a 24-hour collection period during Summer Solstice. The NonSorted Genetic Algorithm II (NSGA-II) heuristic method was used with Multi-Objective Optimization on five trials
to advance 25,000 architectures and identify those with maximal coverage, maximal average capacity, and minimal
average latency. 17 architectural choices were identified and, after analysis, five distinct AN design options were
presented to the SPOs decision-maker based on a balance of capability and managerial factors. The methodology
lays a foundation for assessing future AN options given a set of desirable measures.
Date of Conference: September 15-18, 2020
Track: SSA/SDA