Ukraine funding

U.S. coordinator: Carnegie Mellon University
U.S. Principal Investigator: Samuel Pagliarini
Estonian partner: Tallinn University of Technology, Department of Computer Systems
Estonian Principal Investigator: Levent Aksoy
Ukrainian partner: V. N. Karazin Kharkiv National University

Ukraine has its standardized cryptographic algorithms, namely Kalyna and Kupyna, which are significantly dissimilar to other standardized solutions adopted by NIST in the US. For several reasons, including performance and security, hardware implementations of cryptographic algorithms are sought. In this project, the partners based in US, Estonia, and Ukraine will jointly investigate how to implement these algorithms in an Application Specific Integrated Circuit (ASIC) while disseminating knowledge in both directions.

The Estonian partner is responsible for accomplishing three tasks: (i) security-aware design exploration of cryptographic primitives in Kalyna and Kupyna, such as substitution and linear transformation operations; (ii) realization of design architectures targeting high performance and low power dissipation; (iii) secure design of Kalyna and Kupyna against well-known attacks, such as side-channel analysis and fault injection.

Ukraine does not enjoy access to trusted chip fabrication technology. It is imperative that any hardware implementation of the national cryptographic standards should withstand tampering at fabrication and deployment time. This project addresses this gap by sharing ASIC design skills with Ukrainian researchers while sharing best practices related to security closure.

We strive to have Ukrainian universities teaching, in the near future, both software and hardware cryptographic engineering. The latter is not adequately covered by any university in the country as today. To achieve this goal, the chip design tasks executed in this project will be highly reproducible. Design resources will be openly shared on public repositories, including Verilog source files. Synthesis scripts for state-of-the-art commercial chip design tools will also be shared. Finally, the entire design process will be documented through a series of tutorial-style videos that will showcase the entire chip design step by step, from RTL to layout.

U.S. coordinator: Rochester Institute of Technology
U.S. Principal Investigator: Sergey E Lyshevski
Estonian partner: Tallinn University of Technology, Department of Computer Systems
Estonian Principal Investigator: Gert Jervan
Estonian Co-Principal Investigator: Mairo Leier
Ukrainian partner: The National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute"

The overall objective of the project is to design and implement a content-aware concept of object recognition for nefarious conditions, low data quality and uncertain conditions. Different mathematical, signal processing and image processing methods with integrated Machine Learning are combined to increase an object detection and classification accuracy, as well as to improve the robustness and speed.

The role of Tallinn University of Technology is to reduce the complexity of the existing Machine Learning, implement these models on a special accelerator hardware to ensure high accuracy data acquisition with near real-time capabilities on edge computing devices.

Developed methods can be validated in Ukraine on the real devices that usually operate in harsh environments, under physical impacts, line-of-sight scene obstructions, as well as electromagnetic spectrum and interference.

U.S. coordinator: University of Florida
U.S. Principal Investigator: Rafael Munoz-Carpena
Estonian partner: Tallinn University of Technology, Department of Electrical Power Engineering and Mechatronics
Estonian Principal Investigator: Andrii Chub
Ukrainian partners: G.E. Pukhov Institute for Modelling in Energy Engineering (PIMEE), Cyber Diia Platform, National University of Life and Environmental Science of Ukraine jt
Polish partner: Institute of Theoretical and Applied Informatics of Polish Academy of Science
Lithuanian partner: Mykolas Romeris University

Assessing resilience of critical systems is becoming a standard practice to better understand how they perform in critical situations, like pandemics, abnormal climate conditions, or war. This project will develop a new framework of stress-testing based on quantitative analysis of how systems recover under shocks. This project unites researchers from USA, Poland, Lithuania, and Estonia to apply novel analysis methods to critical digital and power infrastructure systems of Ukraine. This project will provide Ukraine with tools for enhancing resiliency of its critical infrastructure systems and catalyze the integration of Ukrainian researchers into the global research community. The Estonian partner will develop highly-resilient last-mile electrification tools allowing uninterrupted electricity supply to critical, sometimes life-sustaining, consumers, like hospitals, command centers, etc.

U.S. coordinator: University of California
U.S. Principal Investigator: Ilya Zaslavsky
U.S. partner: New Mexico State University
Estonian partner: University of Tartu, Institute of Ecology and Earth Sciences (involving Geological Survey of Estonia)
Estonian Principal Investigator: Argo Jõeleht
Ukrainian partners: Ukrainian Hydrometeorological Institute, Taras Shevchenko National University of Kyiv
Polish partners: Polish Geological Institute, Space Research Center of the Polish Academy of Sciences
Lithuanian partner: Vilnius University
Latvian partner: University of Latvia

The project aims to develop solutions to the persistent problem of accurately assessing groundwater, which is the main source of drinking water in many regions, including transboundary areas. Our hypothesis is that integrating hydrogeologic models with satellite and ground-based data to increase the resolution of remote sensing information will allow us to create highly detailed predictions of changes in GW storage and flows across borders. The challenge is that satellite gravity data from GRACE-FO have coarse spatial and temporal resolution and often inaccurate to separate the groundwater signal from other water balance components such as precipitation, runoff, evapotranspiration, snow storage etc. and may limit its utility for analysis of aquifer depletion and resilience at finer scales. The highest quality data typically comes from expensive, sparsely distributed in-situ monitoring. The project combines comprehensive AI-assisted modeling of surface water and groundwater interactions to assess groundwater storage, flows, and aquifer resilience under different scenarios, to support decision-making given incomplete, imprecise, and differently structured in-situ data. There are partners from six countries: the United States, Ukraine, Poland, Latvia, Lithuania, and Estonia. Estonian partners (University of Tartu and Geological Survey of Estonia) are mainly involved with collection of observational data and provide hydrogeological modelling expertise that are needed for ground truthing of developed models, but contribute also to other tasks. The project results will be especially useful to Ukraine, where limitations of the ground observation network require more efficient assessment techniques based on remote sensing.