Exploring HPC and ML Techniques on Various Classical, Machine and Modern Ciphers
Venu Nalla, Mandeep Kumar, V Samhith, G Padmavathi · Indian Journal of Science and Technology · 2025
Objectives: To investigate High Performance Computing (HPC) techniques for communication ciphers. To explore machine cipher classification using encrypted image data and ML for 4G & 5G cipher classification based on ciphertext analysis. Methods: Employed MPI and CUDA + MPI models to assess the strength of ciphers, classified machine ciphers with image data, and applied ML for identification communication cipher classification. Findings: Communication ciphers are resistant to exhaustive key search attacks using HPC. Applied various classification and clustering models on machine ciphers using encrypted image data collected from online, and got the test accuracy up to 93.9%. ML methodologies for classification of 4G and 5G mobile communication ciphers achieved a test accuracy up to 79.2%. Novelty: Strength analysis of various Cryptographic algorithms belonging to Classical, Machine & Mobile communication. Classification of Some machine ciphers & Mobile communication ciphers using HPC solely based on ciphertext analysis contributing novel insights into communication cipher security. Keywords: High Performance Computing, Clustering, Classification, Mobile Communication Ciphers