An improved differential fault analysis attack to AES using reduced searching space

Abstract

Differential Power Analysis against AES proved to be effective, with use mask techniques, we can truncate the relevance and defense DPA attacks successfully. In this paper, we introduce a new mean called DFA (Differential Fault Attack), DFA has been shown successfully to attack AES algorithm with masking. We inject a fault to the intermediate results and other general form to obtain the ciphertext with faults, using the ciphertext we can recover the keys. Firstly, we construct the Sbox/InvSbox distribution tables to make the results sets space is less than 28 with two correct/fault ciphertexts pairs, even, in most cases, the set space are strict in 22. Secondly, we also demonstrated that the model of DFA we constructed can ignore the masking techniques and propose some methods to recover the keys with less time consuming and reduced searching space. Lastly, we make the module into an application and prove the module is effective. © Springer International Publishing Switzerland 2013.