A Novel Chaos-Based Encryption Technique with Parallel Processing Using CUDA for Mobile Powerful GPU Control Center
Harun Emre Kıran
Abstract
Chaotic systems possess unique properties that can be leveraged for cybersecurity. These properties stem from the complex and unpredictable nature of images, which makes it challenging for systems to interpret them. When combined with CUDA, chaotic systems benefit from high-efficiency parallel processing capabilities, allowing for the rapid and secure handling of large data sets. Therefore, chaotic systems can be effectively used to securely store and conceal images. In this study, a CUDA-supported chaos-based parallel processing encryption mechanism for mobile control centers is developed. Encryption processes leverage the powerful GPU of the control center. This allows for the fast encryption and decryption of image data received from multiple devices connected to the control center via wired or wireless connections. For encryption, the Logistic Map is used to generate random numbers. Using this map, image data is subjected to XOR operations, encrypting the R, G, B, and Gray Scale channels of the images. Initially, an analysis of the numbers generated from this map is conducted, followed by a detailed explanation of the encryption technique. The technique is then applied to image data, and image analyses are performed. Finally, the performance of the encryption technique is compared with other studies, and encryption speeds are examined. The results show that the new encryption technique provides significantly fast encryption and security levels comparable to other studies. The key discovery of this research is that the devised mechanism is well-suited for parallel processing, allowing for rapid image encryption using the proposed method. For encrypting large IoT files, random number generation is initially performed, followed by statistical tests. Subsequently, encryption is executed using the developed algorithm, and security analyses are conducted. The performance of the proposed mechanism is compared with other studies in the literature, and the results from image analysis and encryption performance demonstrate that the developed mechanism can be effectively used with high security for IoT applications.