CISAF: A Framework for Risk Assessment of Academic and Research Cyberinfrastructure

arXiv:2509.00266 · cs.CR · Submitted 2025-08-29 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Security Radio. Generated commentary on the latest security and cryptography papers.

Nadia: We're looking at the paper: "CISAF: A Framework for Risk Assessment of Academic and Research Cyberinfrastructure".

Elias: Here's the gist of the paper. Academic and research cyberinfrastructures (AR-CIs) present unique security challenges due to their collaborative nature, heterogeneous components, and the lack of practical security assessment frameworks tailored to their needs CISAF is proposed as a simple, systematic, mission-centric approach to analyze the security posture of a CI and prioritize mitigation actions.

Nadia: Framework Overview CISAF guides administrators through a topdown process: (1) defining unacceptable losses, (2) identifying associated system hazards and critical assets, (3) analyzing possible attack paths that target these critical assets, and (4) analyzing security mechanisms that lie on these attack paths.

Elias: By combining information about the CI architecture, mission, attack vectors, and security mechanisms, CISAF provides a clear overview of potential security risks and offers valuable information to prioritize mitigation actions.

Nadia: Context and Motivation AR-CIs are complex networks that face specific challenges stemming from sporadic funding and an open user base.

Paper summary: Nadia: To wrap up our look at this paper, we’ve been discussing the CISAF: A Framework for Risk Assessment of Academic and Research Cyberinfrastructure. It’s really about providing a simple, systematic way for these complex research networks to analyze their security posture.

Elias: The core contribution is that it moves beyond just listing threats and instead forces the team to think mission-centrically from the very beginning when they define unacceptable losses.

Nadia: So, in simple terms, it’s a tool that takes the unique challenges of academic research networks—like their open nature and sporadic funding—and gives them a clear, actionable path forward for security improvements.

Priya: It helps them understand exactly what risks matter most to their specific mission goals before they start spending time on deep technical fixes.

Elias: The authors are showing that by combining architecture, mission, attack vectors, and existing security mechanisms in this structured way, you get critical information for prioritizing mitigation actions.

Nadia: It’s about giving those teams a common language to talk about risk so they can select the changes that actually make the biggest difference in their cyber risk reduction.

Priya: And while it acknowledges that it's manual and needs more detail later on, it still provides a solid, tailored structure for getting started with this kind of analysis.

Conclusion: Nadia: So we're wrapping up this look at the CISAF paper by Liang and Mirkovic. It’s really about giving these research networks a simple, systematic way to figure out their security posture before they get overwhelmed.

Elias: Yeah, it’s proposing this framework called CISAF. It guides them through four steps: defining what losses are unacceptable, finding the system hazards and the critical assets involved, then tracing the attack paths that target those assets, and finally checking which existing security mechanisms are on those paths.

Priya: I see how that structure helps because it forces a focus on their specific mission. They’re not just looking at general IT threats; they’re linking everything back to what the research itself needs to achieve.

Nadia: Exactly. It takes the messy reality of academic cyberinfrastructure—all those open devices and small teams—and makes it manageable by focusing on mission-centric analysis first, which then lets them prioritize what to fix.

Elias: The authors say they’re inspired by things like MISSION CENTRIC and MBCRA, but they’ve tailored it for these research environments specifically, making it actionable for smaller groups.

Priya: It seems like the main thing here is translating complex security concepts into a practical checklist that a team with limited time and expertise can actually use to make tangible security improvements.

Nadia: Right. So what this means for us in the real world is that we finally have something that’s designed *for* these specific, tricky environments, instead of trying to force a generic framework onto them.

Elias: It gives researchers a way to start their risk assessment process without needing a massive budget or an army of security experts right away.

Priya: But the authors themselves point out that because many steps are still manual, there’s still a lot of work and potential blind spots there, which is something we have to keep in mind.

Nadia: True. It’s not some magic fix; it’s a starting point that helps them see what they need to look at next using other tools like ATT andCK or STRIDE for deeper dives on specific parts of the infrastructure.

Elias: So the authors are setting up a baseline, but they acknowledge that real security usually means following up with more detailed analysis once you have this initial mission-centric map.

Priya: It’s a useful tool for getting those initial conversations started about where the biggest vulnerabilities actually lie within a research network.

Nadia: Yeah, it’s basically giving them a clear roadmap to start their security journey instead of just feeling lost in the complexity of their systems.

Elias: And that process, that systematic approach they’ve outlined, is exactly what we need to think about when we talk about how to actually defend these kinds of complex research projects in the next segment.

Information Sciences Institute, University of Southern California

cs.CR

Submitted: 2025-08-29

Updated: 2026-10-31

Comments: Total of 19 pages: formal text 8 pages, LLM usage + citation for 2 pages, an appendix of the LLM prompt for 9 pages. 2 figures. Pending next revision for submission

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 69/100

The gist: Academic and research cyberinfrastructures (AR-CIs) present unique security challenges due to their collaborative nature, heterogeneous components…

Key concepts

Academic and Research Cyberinfrastructure (AR-CI)
These are complex networks integrating hardware, software, services, and data used by academic researchers. They face unique security issues like sporadic funding, open user bases, and diverse devices (like IoT), making standard security assessments difficult.
CISAF Methodology
This is a top-down process for security analysis. It starts by defining unacceptable losses related to the mission, then identifies system hazards (attacker actions), maps attack paths using an attack tree, and finally checks how existing security mechanisms affect those paths.
Mission-Centric Approach
This core idea focuses the security assessment on the CI's primary mission. It means that when choosing which security fixes to implement, the decision is based on which mitigation actions best protect the most critical parts of that mission.
Attack Tree Construction
This involves creating a diagram showing all possible sequences of steps an attacker could take to achieve a specific hazard or loss. CISAF uses this to visualize potential threats and understand the chain of events leading to a security failure.

Terminology

Summary

Academic and research cyberinfrastructures (AR-CIs) present unique security challenges due to their collaborative nature, heterogeneous components, and the lack of practical security assessment frameworks tailored to their needs CISAF is proposed as a simple, systematic, mission-centric approach to analyze the security posture of a CI and prioritize mitigation actions. p. 0

Framework Overview

CISAF guides administrators through a topdown process: (1) defining unacceptable losses, (2) identifying associated system hazards and critical assets, (3) analyzing possible attack paths that target these critical assets, and (4) analyzing security mechanisms that lie on these attack paths. p. 0 By combining information about the CI architecture, mission, attack vectors, and security mechanisms, CISAF provides a clear overview of potential security risks and offers valuable information to prioritize mitigation actions. p. 0 p. 0

Context and Motivation

AR-CIs are complex networks that face specific challenges stemming from sporadic funding and an open user base. AR-CI funding is often small and time-limited, which limits the size and expertise of their DevOps staff. p. 0 Furthermore, ARCIs may include nonstandard, Internet-enabled devices such as Internet of Things (IoT) devices and automated lab equipment. p. 0 The complexity and security risk of an ARCI increase with increased device diversity. p. 0 p. 0 There is a great need to provide practical, customtailored, and easy-to-follow guidelines to AR-CIs on how to assess their security posture. p. 0 p. 0

CISAF Methodology

The CISAF approach starts with a simplified functional architecture of the given AR-CI and a definition of the AR-CI’s mission. It proceeds by defining adverse events that would jeopardize AR-CI mission, called unacceptable losses. p. 2 Next, it defines hazards, which are attacker actions on the AR-CI architecture that may cause one or multiple losses. p. 2 Then, it develops an attack tree consisting of attack paths describing a series of steps that would let an attacker create a hazard. p. 2 Finally, CISAF adds the existing security mechanisms on attack paths to help AR-CI teams reason how these mechanisms disable or weaken the attack paths. p. 2

Application and Prioritization

CISAF complements existing analysis frameworks by being inspired and adopting elements of MISSION CENTRIC, MBCRA, and CAVE, specifically a missioncentric approach for prioritization of mitigation actions and attack tree construction. The framework allows AR-CI teams to prioritize mitigation actions based on the priority of losses that reside at the end of these paths. p. 2 For instance, in the SPHERE Testbed example, infrastructure misuse (L5) is ranked lowest because it does not directly jeopardize SPHERE’s mission of supporting research. p. 3

Limitations and Future Work

The approach has limitations, including the fact that many steps are manual and prone to oversight by analysts. Protection blind spots exist, as the presence of security mechanisms may create a false sense of security, necessitating the use of systematic models like STRIDE or ATT&CK. p. 6 Furthermore, detection and real-time mitigation are needed for hazards where preventative measures are insufficient. p. 6 Future research should focus on automating steps such as automated topology mapping and asset identification to ensure comprehensive attack tree generation. p. 6

Conclusion

AR-CIs face unique cybersecurity challenges due to their user population, open nature, and small teams. The proposed framework offers a simple way to analyze the cybersecurity posture of these AR-CIs and has been illustrated using the SPHERE Testbed. p. 6 Future research is expected to build on this work to make it more comprehensive and automated, leading to improved cybersecurity of AR-CIs. p. 6

References

[1] National Science Foundation, “Cyberinfrastructure and Advanced Computing,” https://www.nsf.gov/focus-areas/cyberinfrastructure, July 2024, accessed: 2024-07-11

[3] A. Adams, K. Avila, E. Heymann, M. p. 6 p. 6 Krenz, J. R. Lee, B. Miller, and S. p. 6 p. 6 Peisert, “The State of the Scientific Software World: Findings of the 2021 Trusted CI Software Assurance Annual Challenge Interviews,” https://hdl. p. 6 p. 6handle.net/2022/26799

[4] Joint Task Force Transformation Initiative, “Guide for Conducting Risk Assessments,” https://doi.org/10.6028/nist.sp.800-30r1, NIST, Tech. p. 6 p. 6 Rep NIST SP 800-30r1, 2012

[5] R. Ross and et al. p. 6 p. 6, “Developing Cyber-Resilient Systems: A Systems Security Engineering Approach,” https://doi. p. 6 p. 6org/10.6028/nist.sp.800-160v2r1, NIST, Tech. p. 6 p. 6 Rep NIST SP 800-160v2r1, December 2021

[6] The MITRE Corporation, “MITRE ATT&CK®,” https://attack.mitre.org/, 2025, accessed: 2025-03-27

[8] R. K. De Naray and A. p. 6 M. Buytendyk, “Analysis of Mission Based Cyber Risk Assessments (MBCRAs) Usage in DoD’s Cyber Test & Evaluation,” Naval Postgraduate School, Tech. p. 6 Rep., June 2022

[10] R. Sahay, D. S. Estay, W. Meng, C. D. Jensen, and M. p. 7 B. Barfod, “A comparative risk analysis on CyberShip system with STPA-Sec, STRIDE and CORAS,” Computers & Security, vol. p. 7 128, p. 103179

[11] G. Bakirtzis, B. p. 7 p. 7 Carter, C. Fleming, and C. p. 7 p. 7 Elks, “MISSION AWARE: Evidence-Based, Mission-Centric Cybersecurity Analysis,” arXiv preprint arXiv:1712. p. 7 p. 701448, 2017

[12] J. L. Pecharich, A. p. 7 p. 7 Viswanathan, S. p. 7 p. 7 Stathatos, B. p. 7 p. 7 Wright, and K. p. 7 p. 7 Tan, “Mission-centric Cyber Security Assessment of Critical Systems,” in AIAA SPACE 2016, September 2016

[13] C. Jackson, B. p. 7 p. 7 Cowles, S. Russell, E. p. 7 p. 7 K. Adams, R. Kiser, R. Ricks, and A. p. 7 p. 7 Shankar, The Trusted CI Framework Implementation Guide for Research Cyberinfrastructure Operators

[14] NIST, “The NIST Cybersecurity Framework (CSF) 2.0,” https://doi. p. 6org/10.6028/nist.cswp.29, Tech. Rep. NIST.CSWP.29, February 2024

[15] SPHERE, “Collaborate - SPHERE,” https://sphereproject.net/collaborate/, 2025, accessed: 2025-08-07

[16] FABRIC, “FABRIC Knowledge Base,” https://learn.fabrictestbed. p. 7 p. 7net/knowledge-base/creating-or-joining-a-project/, May 2023, accessed: 2023-05-31

[17] ACCESS, “Access Allocations Policies,” https://allocations.accessci.org/allocations-policy--eligibility, 2025, accessed: 2025-03-21

[18] Sage, “Edge Apps: Sage,” https://sagecontinuum.org/docs/category/edgeapps, 2025, accessed: 2025-03-20

[19] ZEUS, “Submit a Proposal: Zeus,” https://zeus.engin.umich.edu/users/submit-a-proposal/, 2025, accessed: 2025-03-21

[20] ngEHT, “Opportunities at ngEHT,” https://www.ngeht.org/opportunities, 2025, accessed: 2025-04-08

[21] Center for Neutron Science, “Neutron Spin Echo for the Nation,” http://cns.che.udel.edu/neutron-spin-echo-for-the-nation/, 2025, accessed: 2025-04-08

[1] National Science Foundation, “Cyberinfrastructure and Advanced Computing,” https://www.nsf.gov/focus-areas/cyberinfrastructure, July 2024, accessed: 2024-07-11

[3] A. Adams, K. Avila, E. Heymann, M. Krenz, J. R. Lee, B. Miller, and S. Peisert, “The State of the Scientific Software World: Findings of the 2021 Trusted CI Software Assurance Annual Challenge Interviews,” https://hdl.handle.net/2022/26799

[4] Joint Task Force Transformation Initiative, “Guide for Conducting Risk Assessments,” https://doi.org/10.6028/nist.sp.800-30r1, NIST, Tech. Rep NIST SP 800-30r1, 2012

[5] R. Ross and et al., “Developing Cyber-Resilient Systems: A Systems Security Engineering Approach,” https://doi.org/10.6028/nist.sp.800-160v2r1, NIST, Tech. Rep NIST SP 800-160v2r1, December 2021

[7] M. Howard and S. p. 6 Lipner, The Security Development Lifecycle. p. 6 USA: Microsoft Press, 2006

[9] Young Jr., William, “Intro to STPA for Security,” in Proceedings of the MIT STAMP Workshop, July 2020

[11] G. Bakirtzis, B. Carter, C. Fleming, and C. Elks, “MISSION AWARE: Evidence-Based, Mission-Centric Cybersecurity Analysis,” arXiv preprint arXiv:1712.01448, 2017

[12] J. L. Pecharich, A. Viswanathan, S. Stathatos, B. Wright, and K. Tan, “Mission-centric Cyber Security Assessment of Critical Systems,” in AIAA SPACE 2016, September 2016

[13] C. Jackson, B. Cowles, S. Russell, E. K. Adams, R. Kiser, R. Ricks, and A. Shankar, The Trusted CI Framework Implementation Guide for Research Cyberinfrastructure Operators

[14] NIST, “The NIST Cybersecurity Framework (CSF) 2020,” https://doi.org/10.6028/nist.cswp.29, Tech. Rep. NISTCSWP.29, February 2024

[15] SPHERE, “Collaborate - SPHERE,” https://sphereproject.net/collaborate/, 2025, accessed: 2025-08-07

[16] FABRIC, “FABRIC Knowledge Base,” https://learn.fabrictestbed.net/knowledge-base/creating-or-joining-a-project/, May 2023, accessed: 2023-05-31

[17] ACCESS, “Access Allocations Policies,” https://allocations.accessci.org/allocations--policy--eligibility, 2025, accessed: 2025-03-21

[18] Sage, “Edge Apps: Sage,” https://sagecontinuum.org/docs/category/edgeapps, 2025, accessed: 2025-03-20

[19] ZEUS, “Submit a Proposal: Zeus,” https://zeus.engin.umich.edu/users/submit-a-proposal/, 2025, accessed: 2025-03-21

[20] ngEHT, “Opportunities at ngEHT,” https://www.ngeht.org/opportunities, 2025, accessed: 2025-04-08

[21] Center for Neutron Science, “Neutron Spin Echo for the Nation,” http://cns.che.udel.edu/neutron-spin-echo-for-the-nation/, 2025, accessed: 2025-04-08

--- Page 1 ---

CISAF: A Framework for Estimating the Security Posture of Academic and Research Cyberinfrastructure Qishen Liang Information Sciences Institute University of Southern California Los Angeles, United States samliangsk@gmail.com Jelena Mirkovic Information Sciences Institute University of Southern California Los Angeles, United States mirkovic@isi. p. 0edu Brian Kocoloski Information Sciences Institute University of Southern California Los Angeles, United States briankoco@gmail. p. 0com Abstract—Academic and research cyberinfrastructures (ARCIs) present unique security challenges due to their collaborative nature, heterogeneous components, and the lack of practical security assessment frameworks tailored to their needs We propose Cyber Infrastructure Security Analysis Framework (CISAF) – a simple, systematic, mission-centric approach to analyze the security posture of a CI and prioritize mitigation actions. p. 0 p. 0 CISAF guides administrators through a topdown process: (1) defining unacceptable losses, (2) identifying associated system hazards and critical assets, (3) analyzing possible attack paths that target these critical assets, and (4) analyzing security mechanisms that lie on these attack paths. By combining information about the CI architecture, mission, attack vectors, and security mechanisms, CISAF provides a clear overview of potential security risks and offers valuable information to prioritize mitigation actions. Index Terms—Internet security, Control system security, Information systems, Systems modeling, Threat modeling. p. 0

--- Page 2 ---

  1. Introduction Academic and research cyberinfrastructures (AR-CIs, for short) are complex networks, with integrated hardware, software, services, and datasets. p. 0 Like any networked resource, AR-CIs are exposed to outsider and insider attacks. p. 0 But unlike many commercial networks, AR-CIs face specific challenges that come from sporadic funding and an open nature of their user base. p. 0 AR-CI funding is often small and time-limited, which limits the size and expertise of their DevOps staff and the effort they can invest in security. p. 0 In addition, ARCIs are highly motivated to support various use cases, some of which may create additional security risks. p. 0 Further, ARCIs may include nonstandard, Internet-enabled devices, such as Internet of Things (IoT) devices and automated lab equipment. p. 0 The complexity and security risk of an ARCI increase with increased device diversity. Lastly, AR-CIs often run custom management software, which may not be sufficiently secure. p. 0 There is a great need to provide practical, customtailored, and easy-to-follow guidelines to AR-CIs on how to assess their security posture. Among the existing cybersecurity frameworks for IT and OT systems, NIST’s standard is one of the most widely accepted security guidelines. p. 0 NIST SP 800-30 [4] and NIST SP 800-160 [5] both provide cybersecurity frameworks for designing Internet-enabled systems and infrastructures. p. 0 Both documents provide highlevel analytic practices and processes that are applicable to IT and OT systems, but they are not immediately and directly applicable to AR-CIs. p. 0 ATT&CK [6] and STRIDE [7] frameworks provide structured approaches to enumerate threats for single devices or services, but they cannot be readily extrapolated to an AR-CI’s scale and complexity. p. 0 Mission Based Cyber Risk Assessments (MBCRAs) [8], System-Theoretic Process Analysis for Security (STPA-Sec) [9], [10], and MISSION AWARE [11] provide frameworks for mission-based security analysis for Cyber-Physical Systems (CPS), where the computational elements are tightly integrated with and control physical processes, while Cyber Analysis Visualization Environment (CAVE) [12] provides one for mission-critical systems, where any failure can lead to catastrophic consequences. p. 0 These security frameworks cannot be directly applied to all elements of AR-CI, and they are prohibitively time-consuming to employ, given the size of AR-CI teams and funding. p. 0 We propose Cyber Infrastructure Security Analysis Framework (CISAF) – a simple, systematic, mission-centric approach to analyze the security posture of a CI and prioritize mitigation actions. CISAF makes the following contributions: Customized for Each AR-CI. p. 0 AR-CIs are designed and constructed by different teams and organizations; they serve different purposes, and thus have different security challenges. p. 0 Easy to use and actionable. p. 0 CISAF assessment is easy for AR-CI teams to understand and apply. p. 0 It produces actionable results, which can lead to tangible improvements in AR-CI security posture. p. 0 Allows for Prioritization of Mitigation. p. 1 CISAF analysis provides critical information to analysts to reason about different cybersecurity improvements, and select those that make the biggest reductions in their cyber risk. p. 1 CISAF complements existing analysis frameworks. p. 1 It is inspired and adopts elements of MISSION CENTRIC, MBCRA, and CAVE [8], [11], [12], specifically a missioncentric approach for prioritization of mitigation actions and attack tree construction. p. 1 AR-CI teams could follow CISAF up with more in-depth analysis, using other approaches, such as ATT&CK and STRIDE, on select pieces of their infrastructure that CISAF highlighted as needing improvements. p. 1

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improve risk prioritization by using CISAF to prioritize mitigation actions based on losses residing at different levels of priority, such as prioritizing paths leading to hazards associated with high-priority losses like experimental data leak (L1).

  2. Enhance system resilience by focusing mitigation efforts on attack paths that have few or no security mechanisms, as these are identified as candidates for mitigation actions.

  3. Develop real-time detection and mitigation capabilities for insider attacks or infrastructure misuse by complementing preventative measures with detection and real-time mitigation approaches where some attack paths have none or very few security mechanisms in place.

  4. Improve security posture assessment by automating steps such as automated topology mapping, asset identification, vulnerability scanning to ensure comprehensive attack trees are generated and no paths are missed.

Sources

Related papers