Fault Injection Resilience

Fault Injection Resilience,10.1109/FDTC.2010.15,Sylvain Guilley,Laurent Sauvage,Jean-Luc Danger,Nidhal Selmane

Fault Injection Resilience   (Citations: 2)
BibTex | RIS | RefWorks Download
Fault injections constitute a major threat to the security of embedded systems. Errors occurring in the cryptographic algorithms have been shown to be extremely dangerous, since powerful attacks can exploit few of them to recover the full secrets. Most of the resistance techniques to perturbation attacks have relied so far on the detection of faults. We present in this paper another strategy, based on the resilience against fault attacks. The core idea is to allow an erroneous result to be outputted, but with the assurance that this faulty information conveys no information about the secrets concealed in the chip. We first underline the benefits of FIR: false positive are never raised, secrets are not erased uselessly in case of uncompromising faults injections, which increases the card lifespan if the fault is natural and not malevolent, and FIR enables a high potential of resistance even in the context of multiple faults. Then we illustrate two families of fault injection resilience (FIR) schemes suitable for symmetric encryption. The first family is a protocol-level scheme that can be formally proved resilient. The second family mobilizes a special logic-level architecture of the cryptographic module. We notably detail how a countermeasure of this later family, namely dual-rail with precharge logic style, can both protect both against active and passive attacks, thereby bringing a combined global protection of the device. The cost of this logic is evaluated as lower than detection schemes. Finally, we also give some ideas about the modalities of adjunction of FIR to some certification schemes.
Cumulative Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.
Sort by: