Working Group K

AgendaWorking Groups

Working Group K

Space Situational Awareness: An Ontological Approach

Fees

$150 USD General Workshop Fee includes access to Working Groups

Date

Wednesday – March 2, 2022

Time

11:00 AM – 1:30 PM PT

Description

Background Space-based and space ground system assets (together: space assets) are an integral part of the U.S. defense enterprise. Specifically, U.S. space assets support real-time data transfer and remote communications, and play a vital role for intelligence, surveillance and reconnaissance. The disruption of even a single one of these assets can seriously degrade the entire intelligence gathering and communications networks, resulting in further vulnerabilities of vital infrastructure. Ensuring the safety of critical U.S. space assets is no longer solely about maintaining accurate tracking estimates or a catalog of space-object characteristics. Rather, it is a multidisciplinary effort, which requires real-time computer-tractable data and information about all space assets, their functions, capabilities, status, actions, and interconnections, all of which are necessary to make advances towards being able to provide for proactive protection within this critical arena.

This workshop is in four parts, devoted to:

  1. The new Center for Space Cyber Security and Cyber Strategy – (CS)3 – established in the University at Buffalo
  2. The ecosystem of Space Domain Ontologies developed by CUBRC in Buffalo
  3. An approach to space event characterization in support of risk estimates
  4. The goals of the new Department of Defense and Intelligence Community Ontology Working Group (DIOWG), and its implications for the space and cyber domains.

1. The mission of the Center for Space Cyber Security and Cyber Strategy (CS)3

Ensuring the protection of U.S. space assets requires a complete and accurate understanding of the space operational picture (SOP). Maintaining an accurate SOP is difficult due to the sheer volume of objects as well as the dynamic nature of the space environment. Further complicating realization of the SOP are the large quantities of multi-INT data relating to multiple facets of the SOP. Historically, many of these data sources have existed in isolation, but new efforts including the newly formed Combined Space Operations Center (CSpOC) have sought to fuse a multitude of disparate multi-INT data sources which give analysts and decision makers more information to comprehend and assess the SOP.

One of the most challenging aspects in assessing the SOP is that it is difficult for a human to effectively interface to the plethora of available multi-INT data in order to make decisions in a timely manner. To that end, effective technologies for supporting protection of U.S. space assets must necessarily provide for clear and effective dissemination of complex information to the end users. This is especially vital for space cybersecurity. This domain is growing increasingly important for not only defense satellites but commercial ones as well. For example, SpaceX’s Starlink satellites will need to be reliably cyber-hard to protect data both from the user and in-space perspective.

Current trends on space cybersecurity focus on making sure that satellites are built to mitigate cyberattacks, i.e. cyber-resilient satellites. In essence current cybersecurity treats a satellite computer just like any other computer. But, the space domain requires treatment of other aspects that are traditionally not considered in cybersecurity measures. For example, for low-Earth orbiting satellites, interactions with ground stations are limited to a few minutes at a time because of the high speeds of the satellites, which yields a limited time-window of opportunities for an attack. Alternatively, high-Earth orbiting satellites, such as those found in geostationary orbits, are seen at all times but only from specific locations on Earth’s surface. Therefore, the space domain presents both temporal and spatial obstacles to cybersecurity.

Ground stations, too, must be secure from cyberattacks. The ever-growing number of ground stations for individual missions, including university-based ground station facilities, causes a further level of complexity to the SOP. Yet another concern is the type of satellite targeted. For example, targeting a specific NASA science mission may cause damage to only one asset. But, targeting specific satellites such as the Tracking and Data Relay Satellites, which are used to provide near-constant communication links between the ground and orbiting satellites, such as Hubble and the International Space Station, may cause much wider damage. These are important domain-boundary issues that need to be clearly defined and specified.

The new (CS)3 center is the only academic, system-focused, space cyber center in the United States. Our center encompasses the interdisciplinary capabilities of several departments and institutional partners within the University at Buffalo School of Engineering and Applied Sciences.

2. The Space Domain Ontologies (SDO)

The Space Domain Ontologies have been developed over the course of 6 years by CUBRC Inc. as part of the Common Core Ontologies (CCO) ecosystem. The Space Domain Ontologies (SDO) ecosystem currently contain more than 800 terms specific to the space domain and import an additional 2,200+ terms to provide rich semantics that integrate space with other domains. Space object and ground station data is aligned with classes and relations in the SDO and stored in a dynamically updated Resource Description Framework (RDF) triple store. The RDF can be queried to support Space Situational Awareness (SSA) in general and to address the specific needs of spacecraft and ground system operators and other users. For each type of entity in the domain, the SDO provides a definition, a classification, and relational links to other entities. The ontologies represent the reality and complexity of entities in the space domain.

The SDO provides a single controlled vocabulary which is used to integrate heterogenous data sources and make the unified data available for utilization by any number of applications. The fundamental idea is that, for data to be unambiguous across all applications, it must be properly represented in a comprehensive ontology that leverages a robust upper and mid-level ontology ecosystem.

We will present the architecture and content of the Space Domain Ontologies and show how they can be extended and combined with CUBRC’s Cyber Ontology to create a Space Cybersecurity Ontology. The Space Cybersecurity Ontology will be made available for use by members of the space cybersecurity community across the globe to communicate more efficiently and develop a shared understanding of prominent ideas within the field. The approach is exemplified in the development of the Common Core Ontologies (CCO), a suite of modular ontologies based upon Basic Formal Ontology (BFO).

3. Space Event Characterization for Risk Assessment

CUBRC and University at Buffalo ontologists and data fusion scientists have been developing a method for space event characterization based on the method of conceptual spaces (CS), which combines aspects of a symbolic (rule-based) and an associationist (feature-based) concept representation. A conceptual spaces model is, in effect, a geometric representation of concepts as they are understood by human experts, assigning for instance metrics for similarity and distance between concepts. We extend this geometrical approach to create a representation of concepts relevant to the domain of threats to satellites. This includes concepts relating to features of satellites, kinetic kill events involving satellites, and threat values associated with satellites. The CS model is applied to entities of these types to yield a measure of relative risks of space events under different characterizations.

4. The DoD and Intelligence Community Ontology Working Group (DOIWG)

To conclude, we will outline activities of the new DoD and IC Ontology Working Group, which has been established as part of the Joint Artificial Intelligence Center (JAIC). We will discuss the goals of the Working Group, and provide an account of its connections to related initiatives such as the INCITS mid-level ontology project and the Space Domain Ontologies project referred to above. We will also describe the existing and projected applications of ontology to artificial intelligence which provide the rationale for the foundation of the DOIWG within the JAIC.

Rationale for this Workshop The last year has seen several new developments relevant to the space cybersecurity and ontology domains. This workshop will describe these developments to obtain feedback from members of the GSAW community.

Format The working group meeting will consist of four short (20 minute) presentations, each of which will be followed by discussion. The meeting will conclude with a general discussion relating to potential future developments.

Leads John Crassidis and Barry Smith, University at Buffalo

Biographies

John L. Crassidis is a leading authority on space situational awareness. He has worked at NASA Goddard Space Flight Center on a variety of missions across the entire domain of space science. He is a University at Buffalo Distinguished Professor and the Director of the Center for Multisource Information Fusion, which is the only dedicated data/information fusion center in the country. He is principal author of the textbook Optimal Estimation of Dynamic Systems, and second author of another textbook entitled Fundamentals of Spacecraft Attitude Determination and Control. He is also Founder and Director of the University at Buffalo Nanosatellite Laboratory, which is currently building three satellites, two of which are funded by the Air Force and one which is funded by NASA.

Barry Smith is a prominent researcher in the ontology field, and he plays an influential role in the Department of Defense and Intelligence Community Working Group (DIOWG) recently established by the Joint Artificial Intelligence Center (JAIC). Smith is principal developer of Basic Formal Ontology (BFO), the ISO/IEC 21838 standard top-level ontology, which is the most widely used resource in the ontology field. Since 2004, BFO has been used as backplane for an open systems modular approach to ontologies. This approach is currently being applied in over 400 domain ontology initiatives to create suites of interoperable ontologies for use for example in intelligence analysis and related fields, and including the suite of Space Domain Ontologies developed to support space situational awareness.