Faculty Dana Dachman-Soled
National Science Foundation
This is an eight-year, $495K 2015 NSF CAREER Award.
Cryptography is an essential and widespread tool for providing information security. Traditional cryptographic models assume only black-box, or input/output access to devices, thereby limiting the class of attacks. In practice, an attacker may obtain non-black-box access to a device and attack a particular algorithm or specific implementation such as through timing or fault injection attacks. Exploiting non-black-box access has been a remarkably effective technique for compromising cryptosystems.
This project investigates mechanisms to protect against such non-black box attacks. The research goals of this proposal are: (1) design and analyze new algorithms for secure computation in threat models that capture non-black-box attacks, and (2) provide a better understanding of the effectiveness of non-black-box techniques in the construction of cryptosystems.