Derek Chen, The Aerospace Corporation; Chang Zhang, The Aerospace Corporation; Mark Mendiola, The Aerospace Corporation; Ann Chervenak, The Aerospace Corporation; Alex Gonring, The Aerospace Corporation; Jeffrey Won, The Aerospace Corporation; Scott Bergonzi, The Aerospace Corporation; Vincent Kong, The Aerospace Corporation; Eltefaat Shokri, The Aerospace Corporation
Keywords: orbit analysis, orbit propagation, cloud computing, data analytics platform, space domain awareness
Abstract:
There is a growing need to perform orbit analysis on a large number of space objects automatically and with high fidelity. The number of commercial and government satellites and other space objects that need to be tracked is increasing rapidly. There are now frequent launches of dozens of satellites, whose trajectories and orbits need to be tracked from launch to continuous operation. Ventures such as SpaceX Starlink, Amazon Kuiper, and OneWeb have begun launching hundreds of low earth orbit (LEO) satellites that will be used to provide global internet and communication capabilities. Tracking all these space objects is a daunting challenge, since much of this tracking is currently initiated manually using mature, legacy orbit determination applications that run on dedicated computing resources with limited user access. To meet the need to scale the number of space objects for orbit propagation as well as the need to provide broader access to the user community, we propose to automate and parallelize these applications using modern cloud computing infrastructure.
We present an approach that performs high fidelity orbit analysis at large scale on a cloud. We download LeoLabs state vector data from the Unified Data Library (UDL) and perform orbit propagation on these state vectors by running multiple instances of the Aerospace Corporation’s TRACE Trajectory Analysis software on a cloud using the Aerospace Data Exploitation (DEX) analytics platform [4]. In this presentation, we describe the architecture of the system and explain how this approach can be used to track thousands of space objects automatically and efficiently.
Our work utilizes state vectors generated by LeoLabs, a company that operates multiple radar stations around the globe and tracks thousands of objects in Low Earth Orbit (LEO). LeoLabs publishes observations from their radar stations, state and uncertainty estimations for each tracked object, and detailed sensor information to the Unified Data Library (UDL). The UDL is a repository for space situational awareness data from many sources and is being developed by BlueStaq for the Air Force Research Laboratory (AFRL) and Air Force Space and Missile Systems Center (SMC).
Our orbit propagation is done using TRACE, the Aerospace Corporation’s Trajectory Analysis and Orbit Determination Program. TRACE has been developed and maintained since the 1960s and is used in the design and analysis of problems associated with high-accuracy orbital trajectory modeling, ephemeris and parameter estimation, and related error analysis. TRACE provides high-fidelity orbit propagation, orbit determination, and covariance analysis. The application is currently implemented in FORTRAN with some modules in C. TRACE typically runs on a single computer under the Windows or Linux operating systems.
We run these orbit propagation analyses on the DEX Data Analytics Platform, a reference architecture and prototype implementation for an evolvable, responsive, and cloud-based data exploitation platform. DEX was developed as part of the Enterprise Ground Service (EGS) to support satellite operations and Space Domain Awareness. The platform extracts and delivers actionable information by running configurable, general-purpose and domain-specific analytics on heterogeneous data sets. In this presentation, we will describe the architecture of the DEX Data Analytics Platform and explain how its capabilities were used by the TRACE team to easily package and deploy their software.
We will describe in detail the end-to-end orbit propagation workflow that runs on the DEX cloud-based data analytics platform and supports automated, large scale, parallel analysis of many objects in congested, contested space. This work is sponsored by SMC/ECXC, Space C2 and the Kobayashi Maru Program Office. This work supports program office efforts in the areas of Space Domain Awareness (SDA) and data management.
Date of Conference: September 15-18, 2020
Track: Astrodynamics