Non-uniform distributions in quantitative information-flow
Proceedings of the 6th ACM Symposium on Information, Computer and Communications Security
Automatically deriving information-theoretic bounds for adaptive side-channel attacks
Journal of Computer Security
Asymptotic information leakage under one-try attacks
FOSSACS'11/ETAPS'11 Proceedings of the 14th international conference on Foundations of software science and computational structures: part of the joint European conferences on theory and practice of software
Quantitative information flow, with a view
ESORICS'11 Proceedings of the 16th European conference on Research in computer security
Compiler mitigations for time attacks on modern x86 processors
ACM Transactions on Architecture and Code Optimization (TACO) - HIPEAC Papers
Automatic quantification of cache side-channels
CAV'12 Proceedings of the 24th international conference on Computer Aided Verification
AUTOCRYPT: enabling homomorphic computation on servers to protect sensitive web content
Proceedings of the 2013 ACM SIGSAC conference on Computer & communications security
Protecting sensitive web content from client-side vulnerabilities with CRYPTONS
Proceedings of the 2013 ACM SIGSAC conference on Computer & communications security
Hi-index | 0.00 |
We show that the amount of information about the key that anunknown-message attacker can extract from a deterministicside-channel is bounded from above by |O| log (n+1) bits, wheren is the number of side-channel measurements and O is the set ofpossible observations.We use this bound to derive a novelcountermeasure against timing attacks, where the strength of thesecurity guarantee can be freely traded for the resultingperformance penalty. We give algorithms that efficiently andoptimally adjust this trade-off for given constraints on theside-channel leakage or on the efficiency of thecryptosystem. Finally, we perform a case-study that shows thatapplying our countermeasure leads to implementations with minorperformance overhead and formal security guarantees.