Highly secure and low overhead countermeasures against side-channel attacks
Ali Akbar Pammu · 2019
Side-Channel Attack (SCA) is an emerging technique to reveal the secret key of crypto-devices by analyzing the processed-data and their Physical Leakage Information (PLI).The crypto-devices refer to electronic devices such as smartphone, smartcards, wearable devices, and Internet of Things (IoT) implemented with cryptographic algorithms.The processed-data can either be the plaintext or the ciphertext.The PLI includes the power dissipation, electromagnetic interference, acoustic, temperature, and timing information of the crypto-device.The Correlation based Attack (CbA) is the most commonly used SCA.In the CbA, the correlation between the processed-data and the PLI is analyzed to reveal the secret key.In contrast, to secure the secret key against the CbA, countermeasure techniques can be applied to reduce the correlation between the processeddata and the PLI.This thesis pertains to the investigation and design of the highly efficient SCA which is meant to efficiently evaluate the SCA resistance of the proposed countermeasure techniques in the cryptodevice.Thereafter, the proposed countermeasure techniques are implemented with emphases on high resilience against SCA and low power/area overhead.Four main research works have been proposed in this thesis including a highly efficient SCA and three highly secured countermeasures against SCA.The four main research works are explained as follows: First, we proposed a Profiling through Relevance-Learning (PRL) technique on PLI to extract the highly correlated PLI with processed-data, as to achieve a highly efficient yet effective SCA when evaluating the countermeasure techniques.The highly correlated PLI is extracted using the knearest-neighbor, k-NN variance clustering by clustering the high variance sampling points and Plaintext