Audio encryption using <i>n</i> -dimensional chaotic system and genetic algorithm optimized S-Box

Chaotic maps inherently possess cryptographically desirable properties, such as pseudo-randomness and sensitivity to initial conditions. This paper proposes an audio block encryption method integrating an n-dimensional chaotic system with a genetic algorithm-optimized S-boxes. First, a novel n-dimensional discrete hyper chaotic map, ND-CMS, is introduced. Lyapunov exponents, phase diagrams and sequence analysis demonstrate hyperchaotic characteristics of ND-CMS. Subsequently, ND-CMS generates pseudo-random sequences and initial S-boxes. An improved multi-objective genetic algorithm is used to optimize and obtain dynamic S-boxes. Furthermore, the proposed audio encryption scheme and adopts a substitution–diffusion structure. Simulations yield two key findings: (1) The optimized S-boxes exhibit excellent performance, characterized by high nonlinearity and low differential uniformity. (2) The encryption scheme achieves effective audio encryption. Comparative analysis demonstrates superior or competitive performance against existing literature in terms of unified average changing intensity, normalized cross-correlation (NCC), peak signal-to-noise ratio (PSNR) and information entropy, confirming its suitability for secure audio encryption.