Robert Carden, MITRE; Dustin Burchett, MITRE; Harvey Reed, MITRE
Keywords: Space Traffic Management (STM), Space Surveillance Network (SSN), Space Domain Awareness (SDA), Persistence, SNARE, BESTA, GREAT
Abstract:
Increased volume of space traffic and number of Resident Space Objects (RSOs) strains sensing resources and requires increased coordination and automation of Space Domain Awareness (SDA). The Space Surveillance Network (SSN) currently utilizes a network of over 30 ground-based radars and optical telescopes in addition to on-orbit telescopes to provide observations on a catalog of greater than 20,000 RSOs. These observations help detect, track, identify, and catalog RSOs which informs SDA. Due to orbital perturbations, RSO observations need to be periodically refreshed, else the last RSO Two Line Element Set (TLE) may not represent the current orbit.
Presently, the Space Defense Operations Center (SPADOC) employs the system known as the Special Perturbations (SP) Tasker to help maintain catalog accuracy. The SP Tasker provides a daily (but coarse) schedule to the SSN, which in turn provides information in the form of sensor observations to aid in TLE refreshes. SP Tasker is limited by serval key factors: SP Tasker tasks once per day and hence cannot automatically adapt to events as they unfold, and, the premise of tasking via the SP Tasker is that events are fixed, which is not generally the case.
SNARE (Sensor Network Autonomous Resilient Extensible) is a United States Space Force (USSF) effort which decentralizes the tasking of, and collection from, SSN sensors which report to the Combined Space Operations Center (CSpOC). SNARE seeks to transform the current catalog-keeping functionality of the SP Tasker into a real time positional data stream, enabling each sensor to self-task based on reading trusted, decentralized, information. This informs SDA tactical relevance based on improved timeliness of positional information and positional accuracy.
SNARE is composed of sensing nodes (sensors), sensor interfacing nodes, and information validating nodes. Interfacing nodes measure available, up-to-date data to produce “utility” of collect values for each RSO. This is done primarily using the Local Value Function (LVF), which can rank order every object in the catalog to what is important to collect on now. The LVF takes in account requirements of each RSO coupled with the known, up-to-date situational awareness of each RSO. Important objects generally have higher collection values based on their requirements, and objects that go unseen for too long or have high uncertainty in their positional state have higher collect values at a given point in time. Since the LVF is a continuous, deterministic, function over time, each sensor independently arrives on what needs to be tasked on currently, and what can feasibly be done.
Sensors have the responsibility of taking the current ordered list of objects throughout the day to collect on and then determine (in a locally optimized sense) what should be currently collected on to maximize network utility. Sensors then provide the interfacing nodes with observations which are then validated and processed by the validating nodes. These nodes inevitably create or refresh TLEs and provide other situational awareness–which is up-to-date information that is then available to the network for future collects. Validating nodes also prevent redundant collections between sensors. The CSpOC reads the processed SNARE positional state, including collections as they occur, providing CSpOC a real time positional stream of data to further process and publish.
The SNARE positional data stream in near real time enables the resulting SDA to be tactically relevant, where SNARE monitors and responds to events as they unfold. This tactical relevance is a direct result of decentralizing and automating tasking functions which are presently centrally performed in CSpOC. Trusted data from SNARE enables an information provider and consumer relationship between SNARE and CSpOC.
The SNARE effort has completed initial modeling and prototypes with rationale for improving SDA tactical relevance and is presently transitioning to operational prototype phase.
Key findings are that (a) once decentralized, the independent decentralized sensors can make decisions on their own, given access to trusted shared data across the SNARE network, and (b) the decentralized sensors can coordinate using data sharing and provide an improved emergent information gain (sensors influence each other). This paper describes the results of modeling and prototypes, followed by the approach for data sharing in the SNARE network. For example, the prioritization and collection decision processes are described in detail, providing the reader with an understanding into how certain challenges (e.g., legacy Special Perturbations (SP) Tasker once-per-day tasking), can be addressed with decentralization yielding an improved real time positional stream of data. The paper concludes with considerations for information sharing beyond positional data, enabling higher-order functions, such as the BESTA (Blockchain Enabled Space Traffic Awareness) and GREAT (Global Resilient Extensible Autonomous Trusted systems) research efforts.
Date of Conference: September 14-17, 2021
Track: Dynamic Tasking