Development of a Hybrid Cryptographic Scheme Using a Modified Hill Cipher and the Rail Fence Algorithm for Securing Financial Data Transactions

Abstract

Financial data security is crucial as cybercrime continuously targets digital transactions. Classical ciphers are often considered weak, while modern encryptions like AES and RSA demand high computational resources, making them less suitable for small-scale institutions. This research aims to develop a lightweight hybrid cryptography scheme by combining a modified Hill Cipher and an adaptive Rail Fence Cipher to secure financial data transactions efficiently without burdening memory. The study utilized a systematic laboratory experimental method comprising five main stages: literature review, independent algorithm modification, hybrid scheme design, algorithm coding and simulation using Python, and performance evaluation. The Hill Cipher was modified using Modulo 256 to accommodate complete ASCII characters, while the Rail Fence Cipher applied an adaptive transposition pattern to enhance data diffusion. Computational testing on 2,000 rows of simulated financial transaction datasets demonstrated a 100% data integrity recovery rate with zero data loss. Furthermore, the performance evaluation recorded highly efficient execution speeds, averaging 0.000083 seconds for encryption and 0.000110 seconds for decryption per transaction, successfully fulfilling the criteria for lightweight cryptography. These findings imply that the proposed hybrid scheme effectively mitigates the linearity weakness of classical ciphers while maintaining low latency. In conclusion, this hybrid cryptographic architecture provides a robust, highly efficient data protection mechanism ideal for low-resource environments, including small-to-medium enterprises and lightweight web applications. For further research, it is recommended to evaluate the scheme against extreme cyber-attack simulations, such as avalanche effect tests, and fully integrate it into production-scale web architectures.

GS Cited Analysis