A. Fundamental Research
High level report of result or partial result that helped move security science forward-- In most cases it should point to a "hard problem". These are the most important research accomplishments of the Lablet in the previous quarter.

A Monitoring Fusion and Response Framework to Provide Cyber Resiliency

This quarter we prepared our findings regarding the applicability of metamodeling for submission to a conference. We had previously found that our novel metamodeling approach generalized by using 7 models as test cases. We found found that for each test case that the corresponding trained metamodel did a reasonably good job of emulating its respective model. This quarter, we also tested different architectural variants of the metamodel architecture on each of the test models to determine effective architectures that produced more accurate metamodels. We submitted to QEST2021, and our paper was accepted for publication.

Uncertainty in Security Analysis

We continue to improve the maximum weight minimization technique, which is a rare event simulation technique for estimating the probability of an attack causing a significant loss to the network. Current effort focuses on relaxing the monotonicity and symmetry assumption of the network to improve the applicability of the technique.

We are studying two-level simulation techniques (e.g., [8], [9], [10]) and their potential application to quantile estimation of reliability/reachability distributions.

An Automated Synthesis Framework for Network Security and Resilience

• We published one paper in the IEEE PowerTech and presented one research poster at the ACM SIGSIM-PADS conference in the current quarter. We are also working to address the first-round review comments for our IEEE Transactions of Smart Grid paper.
• We continue to study the interdependence between the power system and the communication network to improve resilience in critical energy infrastructures, which addresses the resilient architecture hard problem. In the current quarter, we conduct comparative evaluation between our optimization method and an existing utility practice on the same test system; we add a dysfunctional switch model to our proposed solution to ensure the system stability under new operational constraints (such as voltage and capacity limits); we also formulate a stochastic optimization model to address communication uncertainties; finally, we received the first-round review comments from the IEEE Transactions on Smart Grid and are revising our paper.
• We continue to develop a simulation-based platform for cyber-physical system resilience and security evaluation, which addresses the resilient architecture and scalability hard problem. In the current quarter, we migrate the virtual time system kernel from Ubuntu 12 to Ubuntu 20 with the new compensation mechanism for I/O activities; we develop a large-scale case study to demonstrate the effect of the compensation mechanism; we also explore the impact of GPU intensive applications on the current virtual time system; finally, we presented a work-in-progress research poster at the ACM SIGSIM-PADS conference in May 2021.
• We develop a general and interpretable framework for analyzing PMU data in real-time, which addresses the resilient architecture and security-metrics-driven evaluation hard problems. The proposed framework enables grid operators to understand changes to the current state and to identify anomalies in the PMU measurement data. In the current quarter, we presented the framework at the IEEE PowerTech conference in June 2021.
• We have developed a design and evaluation framework for a self-driving “service provider infrastructure” that leverages our prior work on verification and synthesis to automatically self-configure to become resilient to attacks. Our initial focus Is on network and container orchestration systems, and our first implementation will target Kubernetes. Our platform leverages AI planning algorithms to synthesize steps the system needs to take to protect itself against incoming attacks from an intelligent adversary.

Resilient Control of Cyber-Physical Systems with Distributed Learning

We are developing safety and security analysis approaches for real-life of autonomous and cyber-physical systems using statistical and machine learning techniques. Our approaches rely on distributed and sample-efficient optimization techniques that have been developed in the context of the Multi-armed bandit problem. We have shown how these optimization algorithms can be used effectively for statistical model checking of markov decision processes and hybrid systems. We have built a suite of benchmarks related to online safety analysis of autonomous and semi-autonomous vehicles. Our initial results are very promising as the data usage and the running time of our algorithms can be several orders of magnitude better than existing model checking approaches such as Storm and Prism. Two PhD students are dedicating their research time to the project and the prototype tool has been made available online.

We developed and organized the GRAIC Autonomous Racing Competition which was co-located with CPSWeek 2021. The live event had more than 80 registered members and 30+ attendees. The software framework has been made available to the community for research.

B. Community Engagement(s)
Research interaction in the community including workshops, seminars, competitions, etc.

Invited lectures: Mitra

• Interfaces for models and data in verification and synthesis. Workshop on Learning and Control, and seminars co-located with CPS-IoTWeek 21, chaired by Rafal Wisniewski and Manuela Bujorianu, May 18, 2021.
• Data requirements for estimation and verification. Simons Institute, Theoretical Foundations of Computer Science Seminar, May 11, 2021.
• Mitra participated in a panel discussion at the AFCEA Ideation and Innovation Virtual Event. March 10, 2021. Panel discussion on state of the art and challenges in implementing autonomy

• Kevin Jin will serve as a panelist in the "Dynamic Data-Driven Application Systems" track at the 2021 INFORMS annual conference.

• Kevin Jin is organizing a track on Dynamic Data-Driven Application Systems for 2021 INFORMS.

Publications

Verification and Parameter Synthesis for Stochastic Systems using  Optimistic Optimization, Negin Musavi, Dawei Sun, Sayan Mitra, Sanjay Shakkottai, and Geir Dullerud, to appear in Proceedings of IEEE Conference on Control Technology and Applications (CCTA), September 2021.

Continuous Integration and Testing for Autonomous Racing Software: An Experience Report from GRAIC, Minghao Jiang, Kristina Miller, Dawei Sun, Zexiang Liu, Yixuan Jia, Arnab Datta, Necmiye Ozay and Sayan Mitra. http://mitras.ece.illinois.edu/research/2021/GRAIC_CI_ICRAWP.pdf Contributed paper in ICRA 21 Workshop on Opportunities and Challenges with Autonomous Racing, 31 May, 2021.

Egocentric abstractions for verification of distributed cyber-physical systems. Sung Woo Jeon and Sayan Mitra. IEEE Workshop on the Internet of Safe Things (SafeThings'21), co-located with Oakland, 2021. Won the Best Paper Award.

NeuReach: Learning Reachability Functions from Simulations, Dawei Sun and Sayan Mitra, in preparation, February 2021.

Policy Optimization for Markovian Jump Linear Quadratic Control: Gradient-Based Methods and Global Convergence

 and Parameter Synthesis for Stochastic Systems using  Optimistic Optimization, Joao Jansch-Porto, Bin Hu, and Geir Dullerud, submitted for review, January 2021.

Verifying Stochastic Hybrid Systems with Temporal Logic Specifications via Model Reduction, Yu Wang, Y., Nima Roohi, Matt West, Mahesh Viswanathan, and Geir Dullerud,  submitted to Transactions on Embedded Computing Sys- tems, May 2021.

Linear Bandit Algorithms with Sublinear Time Complexity, Shuo Yang, Tongzheng Ren, Sanjay Shakkottai, Eric Price, Inderjit Dhillon and Sujay Sanghavi, submitted for review, February 2021.

Related recent and forthcoming publications:

Optimistic Optimization for Statistical Model Checking with Regret Bounds, Negin Musavi, Dawei Sun, Sayan Mitra, Sanjay Shakkottai, and Geir Dullerud, July 2020. Presented at the workshop on Symbolic and Numerical methods for Reasoning about Cyber-Physical Systems. 

Full version available online from https://arxiv.org/abs/1911.01537 HOOVER tool available from: https://github.com/sundw2014/HooVer

Verifying Cyber-Physical Systems: A Path to Safe Autonomy, Sayan Mitra, Published by MIT Press, February 16, 2021.

Warm Starting Bandits with Side Information from Confounded Data, N. Sharma, S. Basu, K. Shanmugam and S. Shakkottai, arXiv 2002.08405, 2020. Available at: https://arxiv.org/abs/2002.08405

L2-gain Analysis of Periodic, Event-Triggered Control and Self-Triggered Control using Lifting, N. Strijbosch, G.E.Dullerud, A.Teel, M.Heemels”, to appear IEEE Transactions on Automatic Control, 2021.

Christopher Hannon, Deepjyoti Deka, Dong Jin, Marc Vuffray, and Andrey Y. Lokhov. "Real-time Anomaly Detection and Classification in Streaming PMU Data." IEEE PowerTech Conference, June 2021

Abstract: Phasor measurement units (PMUs) are being pervasively deployed in the grid to provide fast-sampled operational data to aid control and decision-making for reliable operation of the electric grid. This work presents a general and interpretable framework for analyzing PMU data in real-time. The proposed framework enables grid operators to understand changes to the current state and to identify anomalies in the PMU measurement data. We first learn an effective dynamical model to describe the current behavior of the system by applying statistical learning tools on the streaming PMU data. Next, we use the probabilistic predictions of our learned model to principally define an efficient anomaly detection tool. Finally, our framework produces real-time classification of the detected anomalies into common occurrence classes. We demonstrate the efficacy of our proposed framework through numerical experiments on real PMU data collected from a transmission operator in the USA.

Hard problem(s) addressed: security-metrics-driven evaluation, design, development, and deployment

C. Educational Advances
Impact to courses or curriculum at your school or elsewhere that indicates an increased training or rigor in security research.

None to report this quarter.