Search

It is now more crucial than ever to provide data security, as the goals and capabilities of eavesdroppers are constantly changing. Therefore, different developers are creating cipher systems that employ various innovative techniques. Standard ciphers, such as DES and AES, use substitution boxes to ensure secure encryption and decryption of data. The substitution box (S-box) is a core module used in modern ciphers to secure data. This study introduces an efficient and straightforward method that leverages a Residue Number System (RNS) to construct an S-box. Furthermore, the AES algorithm uses the generated S-box to encrypt digital images. Parameters such as entropy, NPCR, and UACI effectively measure the security of the proposed method. The results of the performance and comparison study confirm that the proposed S-box outperforms existing methods, establishing it as a strong candidate for cryptographic use in various image security applications. [ABSTRACT FROM AUTHOR]

Edge computing serves as a critical intermediary for secure data exchange between IoT devices and data centres in healthcare, where the protection of sensitive patient data is paramount. This study presents UdesMec, a comprehensive edge computing framework designed for efficient and secure data storage and processing in such contexts. Focusing on the challenge of implementing robust data security models in the presence of limited resources at edge nodes, UdesMec employs a unified encryption approach that is particularly suited for expert knowledge-based prediction of complex critical medical events. Using secret sharing and homomorphic encryption, it ensures the confidentiality and integrity of sensitive patient data transmitted by IoT devices, while enabling efficient computation of prediction algorithms on a cloud server. The experimental evaluation confirms the strong encryption performance and the ease of use of UdesMec, positioning it as a promising solution for the secure and reliable prediction of complex critical medical events in healthcare systems based on edge computing. [ABSTRACT FROM AUTHOR]

Existing Fully Homomorphic Encryption schemes mainly operate on integer data in the finite field. However, in many real-life applications, the data is in floatingpoint number format, and the homomorphic encryption of floating-point information is challenging. This paper presents a new homomorphic encryption method for matrices whose elements are floating-point numbers. The proposed homomorphic encryption of Floating-Point Data Matrices is applied for secure image processing where the loss of accuracy due to the finite length of the floating-point format representation is insignificant. Here, the encryption/decryption keys are also in floating-point format. New techniques of homomorphic weighted addition, multiplication, and DCT calculation are presented. Homomorphic proxy re-encryption is the novelty of the proposed method. [ABSTRACT FROM AUTHOR]