Monitoring, Fusion, and Response for Cyber Resilience - January 2021
PI: William Sanders
Researchers: Michael Rausch
Special Note: The researchers who previously worked on this project (Mohammad Noureddine, Uttam Thakore, and Ben Ujcich) graduated and are no longer working on the project. Michael Rausch started working on this project this quarter. He had previously worked on the project titled "A Human-Agent-Focused Approach to Security Modeling.
HARD PROBLEM(S) ADDRESSED
This refers to Hard Problems, released November 2012.
- HARD PROBLEM(S) ADDRESSED
Accounting for Human Behavior - Recognizing the influence of human actions on security outcomes, the aim of this project is to make fundamental advances in scientifically-motivated techniques to aid risk assessment for computer security through the development of a general-purpose, easy-to-use formalism that allows for realistic modeling of cyber systems and all human agents that interact with the system, including adversaries, defenders, and users, with the ultimate goal of generating quantitative results that will help system architects make better design decisions.
Our hypothesis is that models that incorporate all human agents who interact with the system will produce insightful metrics. System architects can leverage the results to build more resilient systems that are able to achieve their mission objectives despite attacks. We are particularly interested in performing uncertainty quantification and sensitivity analysis of cyber security models by using specially constructed metamodels to validate cyber security models.
PUBLICATIONS
Papers written as a result of your research from the current quarter only.
- M. Rausch and W.H. Sanders. Stacked Metamodels for Sensitivity Analysis and Uncertainty Quantification of AMI Models. Proceedings of the IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids (SmartGridComm), 2020.
Abstract: Models can help architects design effective and secure advanced metering infrastructure (AMI) deployments. Because of the complex interactions among the numerous smart meters, smart home devices, customers, the utility, and potential adversaries, the models are often complex and have long execution times. In addition, the models often contain a large number of uncertain input variables. Modelers seek to understand the impact of uncertain input variables on the model through the use of sensitivity analysis (SA) and uncertainty quantification (UQ). However, long-running models are not amenable to such techniques, since they require that the model be run many times. One approach to help overcome this challenge is to build a metamodel (a model of the model) that accurately emulates the original model but is much faster. In this paper, we explain an approach we developed to do fast and thorough SA and UQ using a specially designed metamodel of stacked regressors that can be used to analyze AMI models. We demonstrated the approach by applying it to a complex AMI security model. We show that our metamodel is substantially faster than the base AMI model, more accurate than other existing metamodel approaches, and amenable to SA and UQ.
KEY HIGHLIGHTS
Each effort should submit one or two specific highlights. Each item should include a paragraph or two along with a citation if available. Write as if for the general reader of IEEE S&P.
The purpose of the highlights is to give our immediate sponsors a body of evidence that the funding they are providing (in the framework of the SoS lablet model) is delivering results that "more than justify" the investment they are making.
This quarter, we presented our work titled Smart Metamodels for Sensitivity Analysis and Uncertainty Quantification of AMI Models at SmartGridComm and published it in the conference proceedings. We received valuable feedback during the conference and gave people the opportunity to learn more about this novel technique for indirect SA and UQ. Furthermore, one person who attended the conference gave us a high-quality quantitative security model which we intend to use as a test case for our method. We are continuing our efforts to investigate whether an adaptive sampling query-by-committee approach can improve the construction of metamodels which can be used to perform faster sensitivity analysis and uncertainty quantification.
COMMUNITY ENGAGEMENTS
No community engagements this quarter.
EDUCATIONAL ADVANCES:
None to report.