Using PKPSs with Tamper-Responsive Modules

Abstract

The fact that secrets assigned to need to be stored, and computations using them performed, inside a tamper-responsive environment places some additional constraints. One significant problem in the implementation of tamper-responsive modules stems from remnance in volatile memory regions which permits attackers to expose secrets from scavenging foot-prints left behind. A simple security policy — the decrypt only when necessary (DOWN) policy can limit the number of secrets exposed from such an attack. The DOWN policy obviates the need to use inherently vulnerable, and more expensive, multi-step countermeasures as a response to suspected attacks on the integrity of trustworthy modules. Using PKPSs with the DOWN policy, and the DOWN assurance that not more than one secret can be exposed by tampering with a module, has an effect both on the overhead and the security of PKPSs. This chapter provides an in-depth discussion of the effect of using the DOWN policy. The Chapter also lays down algorithms to be executed description inside tamper-responsive boundaries for computing pairwise secrets using different approaches. A more generalized form of the atomic relay algorithm that leverages algorithms for computing pairwise secrets is also provided.