CRYPTO '99 Proceedings of the 19th Annual International Cryptology Conference on Advances in Cryptology
Differential Fault Analysis of Secret Key Cryptosystems
CRYPTO '97 Proceedings of the 17th Annual International Cryptology Conference on Advances in Cryptology
Using Memory Errors to Attack a Virtual Machine
SP '03 Proceedings of the 2003 IEEE Symposium on Security and Privacy
Tamper resistance: a cautionary note
WOEC'96 Proceedings of the 2nd conference on Proceedings of the Second USENIX Workshop on Electronic Commerce - Volume 2
Lest we remember: cold boot attacks on encryption keys
SS'08 Proceedings of the 17th conference on Security symposium
| Hi-index | 0.00 |
There is a long history of security attacks that succeed by violating the system designer's assumptions about how things work. Even if a designer does everything right--within the "obvious" model--such attacks can still succeed. How can we, as designers and verifiers of systems, cope with these "outside-the-box" attacks?The classic examples of assumption-violating attacks are the timing attacks on cryptosystems first introduced by Kocher [1]. Cryptosystems are designed so that an attacker who has black-box access to an implementation (and does not know the secret key) cannot deduce the key. Extensive mathematical analysis of the input-output behavior of cryptographic functions led to the belief (though unfortunately not proof) that an attacker who can observe the input-output behavior of cryptosystems cannot feasibly find the secret key. Kocher showed that even if this is true, the running time of common cryptographic algorithms does depend on the secret key. Though the dependence of running time on the key is complex, Kocher showed how to use randomized experiments to extract enough signal to deduce the key, at least in principle. Brumley and Boneh later showed that such attacks are practical, even across a network[2].