Preventing cryptographic attacks used on the Internet of Things

Abstract

Cryptographic attacks on Internet of Things (IoT) devices are not highly con sidered by the users of IoT. Most cryptographic algorithms commonly used on IoT devices are vulnerable to cryptographic attacks. Cryptography attacks refer to mathematical procedures to crack the secret key of the algorithm used on IoT devices. More needs to be done to prevent attacks on cryptographic algorithms used on IoT devices. The objectives of this study are: (i)To use the Khumbelo Difference Muthavhine (KDM) function to prevent Differen tial Cryptanalysis (DC) attacks in the AES algorithm used on IoT devices. (ii) To apply the Blocker function to prevent Differential-Linear Cryptanal ysis (DL) attacks in the Serpent algorithm used on IoT devices. (iii) To use the Khumbelo function to prevent Linear Cryptanalysis (LC), DC, DL, boomerang, truncated differential, meet-in-the-middle, and zero-correlation linear-distinguisher attacks in the Camellia algorithm used on IoT devices. (iv) Applying the Khumbelo function to protect IoT against LC, DC, DL, boomerang, truncated differential, meet-in-the-middle, and zero-correlation linear-distinguisher attacks. The KDM, Khumbelo, and Blocker functions prevented cryptographic attacks since all 8 x 8 S-Boxes were changed to 8 x 32 S-Box depending on the particular chapter. The analysis produced re markable results in preventing cryptographic attacks from IoT devices using the KDM, Khumbelo, and Blocker functions. The objectives of the study was to block the construction of distiguishers. Distinguishers are used by intrud ers as first step to conduct any cryptographic attack. Once the construction of distiguishers are blocked, therefore no attacks would be established. The study managed to block construction of distiguishers to 0% probability com pared to (i)50% of LC attacks, (ii) 50% of DL attacks, and (iii) 50% of DC attacks. The 8 x 32 S-Box was expected to build distinguishers from 2 8 x 232 = 256 x 4, 294, 967, 296 matrices with 1, 099, 511, 627, 776 elements. Due to computational space, an ordinary computer could not compute 256 x 4, 294, 967, 296 matrices.