Paper 2018/1181 Original: eprint.iacr.org/2018/1181 ↗

Horizontal DEMA Attack as the Criterion to Select the Best Suitable EM Probe

Abstract

Implementing cryptographic algorithms in a tamper resistant way is an extremely complex task as the algorithm used and the target platform have a significant impact on the potential leakage of the implementation. In addition the quality of the tools used for the attacks is of importance. In order to evaluate the resistance of a certain design against electromagnetic emanation attacks – as a highly relevant type of attacks – we discuss the quality of different electromagnetic (EM) probes as attack tools. In this paper we propose to use the results of horizontal attacks for comparison of measurement setup and for determining the best suitable instruments for measurements. We performed horizontal differential electromagnetic analysis (DEMA) attacks against our ECC design that is an im-plementation of the Montgomery kP algorithm for the NIST elliptic curve B-233. We experimented with 7 different EM probes under same conditions: attacked FPGA, design, inputs, measurement point and measurement equipment were the same, excepting EM probes. The used EM probe influences the success rate of performed attack significantly. We used this fact for the comparison of probes and for determining the best suitable one.