IEEE Access (Jan 2020)

Improved SERPENT Algorithm: Design to RGB Image Encryption Implementation

  • Tariq Shah,
  • Tanveer ul Haq,
  • Ghazanfar Farooq

DOI
https://doi.org/10.1109/ACCESS.2020.2978083
Journal volume & issue
Vol. 8
pp. 52609 – 52621

Abstract

Read online

Gigantic data transfer over networks and internet turned in developing a large number of encryption techniques. As a consequence, in parallel, it gave birth to new plans of attacks. In facet of data security, the algorithms specifically Rijndael (Advanced Encryption Standard-AES), SERPENT and TWOFISH are equally considered to be the best. However, they pay less attention in some cryptographic areas like digital image encryption schemes because of their time consumption inferiority. In this paper, we introduce a time diminishing improved version of SERPENT algorithm depending upon chain ring-based substitution boxes (S-boxes) dealing with 8-bit vector instead of 4-bit. The multiplicative substructures in a chain ring have multiple generators and thus the chain ring based S-boxes improve the algebraic complexity of the cipher. Apart from it, the algorithm is utilized in a digital RGB image encryption application where throughout the chain ring operations are performed. The digital image analyses reveal that the proposed scheme consumes less time (i.e. 8.2 microseconds for encryption and 5.8 microseconds for decryption of a 128-bit block) as compared to the existing prominent RGB image encryption techniques. Furthermore, the investigation on the suggested RGB image encryption scheme it is observed that it has a great resistance against the statistical and the differential attacks.

Keywords