Douglas Hope, Georgia Tech Research Institute; Megan Birch, Georgia Tech Research Institute; Christopher E. Cordell, Georgia Tech Research Institute; Keith F. Prussing, Georgia Tech Research Institute; Christopher Valenta, Georgia Tech Research Institute
Keywords: Space Domain Awareness, Space Traffic Management, Network Sensor Designs, Modeling, High Performance Computing, Conjunction Analysis
Abstract:
As the number of satellites and amount of space debris increases, we begin to face the ever more likely scenario of space objects colliding and creating a debris field that potentially causes other collisions, thus initiating a chain reaction. The result of this chain reaction, the Kessler effect, could render large parts orbital regime unusable for space operations and navigation due to the uncontrolled and unknown amount of debris. Preventing this Kessler effect is a significant motivation for developing Space Traffic Management (STM).
The major thrust of STM focuses on conjunction analysis and predicting potential collisions and subsequent mitigation maneuvering of active satellites. Along with the ever-increasing number of satellites in orbit is the current and future proliferation of satellite constellations. Existing and future constellations include Starlink (SpaceX), Kuiper Systems (Amazon), Doves (Planet Labs), PNT systems such as GLONASS (Russia), BeiDou (China), Galileo (Europe), and GPS-III (United States). Given this future, managing space traffic in space will require accurate and timely conjunction analysis and knowledge about the presence and distribution of debris and other space objects in the constellation belt.
Fulfilling this STM requirement will depend on a complex system-of-systems architecture featuring both ground and space-based sensors collaboratively surveying and monitoring different orbital regimes. To address this need in STM, we present the verification and validation (V&V) of a simulation/modeling system, the Dynamic Model Integration and Simulation Engine (DMISE). This V&V process aims to validate the ability of DMISE to render physical radiometric scenes with realistic blurring and noise levels consistent with imagery obtained from a variety of optical-IR sensors. The flexible architecture of DMISE allows it to simultaneously model the orbital dynamics of multiple space objects while rendering a physical radiometric scene of each space object based on the local viewing and illumination geometry for any number of different sensor configurations/locations.
Our verification and validation demonstrate DMISE modeling a wide range of engagement scenarios from wide-field imaging (60+ degrees) to narrow (arc-minutes). In the wide-field imaging regime, we present the ability to model multiple satellite streaks plus background stars and generate large data sets to augment observational data for training machine learning algorithms. We validate DMISE simulated geostationary satellites’ observations over short and long time scales at a more narrow field-of-view using real telescope observations. Based on the verification and validation results, we propose a future path for the use of DMISE in the design of prospective ground and space-based sensors to support data collection for STM and conjunction analysis.
Date of Conference: September 14-17, 2021
Track: Conjunction/RPO