Lightweight image encryption algorithms: design and evaluation

Abstract

In an era dominated by increasing use of multimedia data such as images and videos, ensuring the security and confidentiality of images with real-time encryption is of greatest importance. Traditional encryption algorithms are secure, widely used, and recommended, yet they are not suitable nor computationally efficient for encrypting multimedia data due to the large size and high redundancy inherent in multimedia data. Thus, specialized algorithms for multimedia data encryption are needed. This dissertation explores lightweight image encryption algorithms, specifically designed to address time and resource constraints of realtime image encryption while maintaining the confidentiality and integrity of the multimedia data.

The dissertation classifies image encryption based on the techniques used into seven different approaches and analyzes the strengths and weaknesses of each approach. It subsequently introduces and evaluates three novel algorithms designed to encrypt images with low complexity, high efficiency, and reliable security. These algorithms rely on a combination of permutation, substitution, and pseudorandom keystreams to ensure the security of the encrypted images. The first algorithm is based on chaotic systems. The algorithm is implemented using logistic map, permutations, AES S-box, and a plaintext related SHA-2 hash. The second algorithm is based on Trivium cipher. the algorithm is implemented to work on multi-rounds of encryption using pixel-based row and column permutations, and bit-level substitution. For the third algorithm, the Ascon algorithm selected by the National Institute of Standards and Technology (NIST) to standardize lightweight cryptography applications is evaluated for image encryption.

To evaluate the proposed algorithms, a comprehensive set of security, quality, and efficiency valuation metrics is utilized to assess the proposed algorithms and compare them to contemporary image encryption algorithms.