Towards Finding An Optimal S-box For Lightweight Block Cipher

Abstract

Implementing conventional cryptography like Advance Encryption Standard (AES) requires hardware resources beyond what constrained devices like RFID tags can offer and still perform their primary function. This limitation gave rise to lightweight cryptography to secure constrained devices. The block cipher is the branch of the cryptography scheme that is mostly considered for lightweight cryptography. A key component of the block cipher largely responsible for its security, implementation cost, and efficiency is the Substitution Box (S-box). Most of the time spent in block cipher development is used to find the best S-box with high resistance against known cryptanalysis attacks. However, finding the optimal S-box among the huge possible permutations has always been challenging. The wrong choice of S-box has led to the exploit of some cryptography (cipher). This paper focuses on finding an optimal 4-bit x 4-bit S-box for the lightweight block cipher that will guarantee the cipher security against differential and linear cryptanalysis. We achieved our aim by considering research findings from 1990 to date, to determine the optimal S-box properties and their best values. The S-box properties include and are not limited to differential uniformity, Linearity, and “BOGI Applicability”. Differential uniformity measures resistance to differential attack. S-box Linearity measures resistance to linear cryptanalysis attack. And “BOGI-Applicable S-box” determines if an S-box can implement the “BOGI Strategy”. The “BOGI Strategy” is a strategy that synchronizes the design of a block cipher permutation layer with its S-box to eliminate the S-box weakness. The concluded best values for the S-box characteristics were incorporated into an algorithm and implemented using the C++ programming language. Sample optimal S-boxes were generated using the suggested metric values. The generated S-boxes comply with the “BOGI strategy”, which eliminates the S-box weaknesses that cryptanalysts would otherwise have exploited.