A TECHINQUE FOR IMPROVING THE RELIABILITY OF ENCRYPTION KEYS BASED ON GAUSSIAN DISTRIBUTION

Abstract

This article proposes a technique for generating encryption keys aimed at increasing their reliability and resistance to cryptanalytic attacks. The specific feature of the proposed technique is the use of Gaussian distribution properties in the generation of initial random sequences, after which these sequences undergo specialized processing to generate encryption keys. This approach is designed to increase the entropy of the input data and reduce predictability compared to a number of traditional techniques, while maintaining a satisfactory level of performance for practical applications.

The article provides a theoretical justification for using the Gaussian distribution as a source of variability and describes the general principles of transformations applied to generated sequences in order to increase the reliability and unpredictability of them. A comparative analysis was also conducted, including various assessments such as entropy values, predictability, performance, and complexity of practical implementation. The analysis noted an increase in randomness characteristics and a decrease in the predictability of key material while maintaining acceptable generation time when using this method.

The proposed technique can be used in cryptographic systems where high-quality randomness of keys, limited computing resources, and stable performance are simultaneously required. The research results confirm the promising application of the approach in both software and hardware implementations, and also determine the directions for further work — extended testing on real platforms, deeper cryptanalysis, and standardized evaluation in accordance with modern security requirements.