Signal-to-noise ratio measurements of side-channel traces for establishing low-cost countermeasure design

Improving the countermeasures against side-channel attacks (SCAs) increases the cost of both designing the countermeasures and evaluating SCA resistance. This may force cryptographic ICs to remain vulnerable. The increased cost is due to an indispensable procedure where a large number of side-channel traces need to be analyzed in order to evaluate the SCA resistance. In this work, a low-cost method to design and evaluate countermeasures using the signal-to-noise ratio (SNR) of side-channel traces as design and evaluation criteria is proposed. The method combines two existing methods: A prediction method of correlation coefficients between side-channel traces and a power model based on the SNR of the side-channel traces, and an estimation method of the number of traces needed to disclose the secret key based on the correlation coefficients. We construct a method to measure the SNR of side-channel traces and validate it for the design and evaluation criteria. In our method, the SNR is first calculated from signal and noise variances extracted from side-channel traces by increasing the number of averaging in side-channel trace measurements, and then the correlation coefficients and the number of traces for key-disclose are estimated on the basis of the calculated SNR. We confirmed that the estimated correlation coefficient and the number of traces for key-disclose were in good agreement with the corresponding measured ones. This result demonstrates that the proposed method can accurately measure the SNR of side-channel traces.