Paper 2016/297

A Unified Metric for Quantifying Information Leakage of Cryptographic Devices under Power Analysis Attacks

Liwei Zhang, A. Adam Ding, Yunsi Fei, and Pei Luo

Abstract

To design effective countermeasures for cryptosystems against side-channel power analysis attacks, the evaluation of the system leakage has to be lightweight and often times at the early stage like on cryptographic algorithm or source code. When real implementations and power leakage measurements are not available, security evaluation has to be through metrics for the information leakage of algorithms. In this work, we propose such a general and unified metric, information leakage amount - ILA. ILA has several distinct advantages over existing metrics. It unifies the measure of information leakage to various attacks: first-order and higher-order DPA and CPA attacks. It works on algorithms with no mask protection or perfect/imperfect masking countermeasure.It is explicitly connected to the success rates of attacks, the ultimate security metric on physical implementations. Therefore, we believe ILA is an accurate indicator of the side-channel security level of the physical system, and can be used during the countermeasure design stage effectively and efficiently for choosing the best countermeasure.