2015•Unpublished venueRequires access

Advanced fault analysis techniques on AES

Kazuo Sakiyama, Takanori Machida, Arisa Matsubara

Open publisher page 0 citations

Abstract

Fault analysis research on symmetric-key cipher has been intensively discussed since differential fault analysis (DFA) was proposed in 1997. Output masking for wrong ciphertexts was believed to be the most effective countermeasure of the DFA attacks. However, fault sensitive analysis (FSA), proposed in 2010, can bypass the output-masking countermeasure. Both DFA and FSA require a strict fault injection control with the same plaintext, which is often difficul to realize under a circumstance where faults are randomly injected, e.g., in the case of electromagnetic (EM) fault injections. Although it requires the distribution of faulty ciphertexts, an extended fault analysis technique called NU-FVA, proposed in 2013, can avoid the hardness of the fault injection control. This article reviews the previous fault attacks and discusses their merits and demerits especially focusing on the power of the NU-FVA attack.

About this research paper

What this paper is about

Fault analysis research on symmetric-key cipher has been intensively discussed since differential fault analysis (DFA) was proposed in 1997. Output masking for wrong ciphertexts was believed to be the most effective countermeasure of the DFA attacks. However, fault sensitive analysis (FSA), proposed in 2010, can bypass the output-masking countermeasure. Both DFA and FSA require a strict fault injection control with the same plaintext, which is often difficul to realize under a circumstance where faults are randomly injected, e.g., in the case of electromagnetic (EM) fault injections. Although it requires the distribution of faulty ciphertexts, an extended fault analysis technique called NU-FVA, proposed in 2013, can avoid the hardness of the fault injection control. This article reviews the previous fault attacks and discusses their merits and demerits especially focusing on the power of the NU-FVA attack.

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Fault analysis research on symmetric-key cipher has been intensively discussed since differential fault analysis (DFA) was proposed in 1997. Output masking for wrong ciphertexts was believed to be the most effective countermeasure of the DFA attacks. However, fault sensitive analysis (FSA), proposed in 2010, can bypass the output-masking countermeasure. Both DFA and FSA require a strict fault injection control with the same plaintext, which is often difficul to realize under a circumstance where faults are randomly injected, e.g., in the case of electromagnetic (EM) fault injections. Although it requires the distribution of faulty ciphertexts, an extended fault analysis technique called NU-FVA, proposed in 2013, can avoid the hardness of the fault injection control. This article reviews the previous fault attacks and discusses their merits and demerits especially focusing on the power of the NU-FVA attack.

Key concepts: Countermeasure, Fault (geology), Computer science, Fault injection, Plaintext, Power analysis, Block cipher, Masking (illustration)

Related papers

Back to paper searchBrowse research topicsOriginal source
Advanced fault analysis techniques on AES — Research Paper | ScholarLens