About Our Research

Praesent quis arcu porttitor, ullamcorper quam sit amet, efficitur ligula. Nullam gravida imperdiet metus sed consequat. Maecenas porta purus et pellentesque sollicitudin. Donec pellentesque risus vitae suscipit imperdiet. Donec ex enim, tincidunt id velit non, scelerisque cursus enim. Proin aliquet bibendum odio, vel efficitur lacus iaculis eget. Integer mattis pharetra elementum. Duis laoreet tempus iaculis. Nunc placerat pretium porta.

Go to top

Ongoing Research

Secure Hardware Verification

Supervisor(s): Leonid Azriel
Research document
Status: available

Security verification is a challenging task, since it requires coverage of a wider scenario space than functional verification. To address the security verification of software, the Common Weaknesses Enumeration taxonomy was introduced, a catalogue of software security issues. Nowadays, a variety of tools exist that try to find weaknesses in software code. Following the software path, a hardware CWE database was added and now includes hundreds of entries.
The question is whether specific methods from the software domain can be used in hardware security analysis. One software verification paradigm that might work for hardware is taint analysis, which is known in the hardware world as information flow tracking. Such analysis may use graph methods to track the information flow between different security domains. Early methods are limited to a small scale, due to the required intimate knowledge of every design detail and intrinsic algorithm limitations.
Two topics are suggested for the research:

Using Graph Neural Networks (GNN) to detect security vulnerabilities
Using Quantitative Information Flow (QIF) for secure hardware verification

Go to top