Real-time photovoltaic voltage data-driven dynamic green s-box generation framework for smart grid security

Abstract

In this study, a dynamic substitution box (s-box) generation method is proposed using real-time voltage data obtained from photovoltaic (PV) panels as an environmentally sensitive entropy source. These data, affected by environmental conditions such as solar radiation and temperature, exhibit irregular and unpredictable temporal behavior, offering strong potential for cryptographic randomness generation. In this context, the 0-1 test, Kolmogorov Entropy (KE), scale-index analysis, and the National Institute of Standards and Technology (NIST) SP 800-22 test suite are applied to PV voltage data collected at 1-minute intervals, experimentally confirming their chaotic dynamics and high-entropy characteristics. Following appropriate preprocessing and nonlinear mathematical transformations, the validated data are utilized to generate 8-bit s-box structures. The cryptographic strength of the generated s-boxes is quantitatively evaluated using standard security metrics, including nonlinearity (NL), Differential Probability (DP), Linear Probability (LP), Strict Avalanche Criterion (SAC), and Bit Independence Criterion (BIC). The findings reveal that PV data can serve as a functional resource not only for energy monitoring but also for randomness-based cryptographic construction. This approach enables lightweight cryptographic solutions for resource-constrained edge devices requiring <1 KB of memory and supports context-aware data security in renewable energy infrastructures.