Search

Until recently, power analysis is one of the most popular research issues among various side channel analyses. Since Differential Power Analysis had been first proposed by Kocher e t al., various practical power analyses correspond with software/hardware cryptographic devices have been proposed. In this paper, we analyze vulnerability of countermeasure against power analysis exploiting single power trace of public cryptogr aphic algorithm. In ICICS 2010, Clavier et al. proposed Horizon tal Correlation Analysis which can recover secret information from a single exponentiation trace and corresponding countermeasures . "Blind operands in LIM", one of their countermeasures, exploits additive blinding in order to prevent leakage of intermediate value related to secret information. However, this countermeasu re has vulnerability of having p ower leakage that is dependant with the message known by an adversary. In this paper, we analy zed vulnerabilities by three attack scenarios and proved them by practical correlation power analysis experiments.Keywords: Side-channel power analysis, horizontal correlation analysis, l ong integer multiplication, additive operand blinding countermeasure

Collision attacks using side channel analysis confirm same inte rmediate value and restore sensitive data of algorithm using this point. In CHES 2011 Clavier and other authors implemented the improved attack using Blacklist so they carried out the attack successfully using less plaintext than before. However t hey did not refer the details of Blacklist method and just performed algorithms with the number of used plaintext. Therefo re in this paper, we propose the specific method to carry out efficient collision attack. At first we define basic concepts, terms, and notations. And using these, we propose various metho ds. Also we describe facts that greatly influence on attack perform ance in priority, and then we try to improve the performance of this attack by analyzing the algorithm and structuring more efficient one. Keywords: Side Channel Attack, Collision Attack, Masked AES I.서 론 * 부채널 공격(Side channel attack)[1]은 디바이스가 암호 알고리즘을 구동하는 동안 발생하는 연산 접수일(2014년 10월 17일), 수정일(2014년 11월 25일), 게재확정일(2014년 11월 25일)* 본 연구는 미래창조과학부 및 정보통신산업진흥원의 대학IT연구센터육성 지원사업의 연구결과로 수행되었음 (NIPA-2014-H0301-14-1004)†주저자, hahehohu3875@korea.ac.kr‡교신저자, shhong@korea.ac.kr(Corresponding author)

Esmbedded devices such as smart cards and electronic passports highly demand security of sensitive data. So, the secure implementation of the cryptographic system against various side -channel attacks are becoming more important. In particular, th e fault injection attack is one of the threats to the cryptosystem and can destroy the whole system only with single pair of the plain and cipher texts. Therefore, the implementors must consid er seriously the attack. Several techniques for preventing fault injection attacks were introduced to a variety of the cryptosys tem, But the countermeasures are still inefficient to be applied to the classical RSA cryptosystem. This paper introduces an efficient countermeasure against the fault injection attack for the classical RSA cryptosystem, which is based on the famous Fermat 's theorem. The proposed countermeasure has the advantage that it has less computational overhead, compared with the previous countermeasures. Keywords: classical RSA, Fault Injection Attack, Fermat's Theorem