Audio Signal Encryption Using Memristor-Based Hyperchaotic Dynamics and Statistical Randomness Validation

Abstract

Abstract. The growing demand for secure multimedia transmission, particularly in real-time and resource-limited environments, has emphasized the need for robust audio encryption techniques. This study proposes a novel encryption method using a hyperchaotic four-dimensional system driven by a memristor-based circuit. The hyperchaotic system, verified through Lyapunov exponents and bifurcation analysis, forms the core of a high-entropy pseudo-random number generator (PRNG). The generated sequences pass all subtests in the NIST SP800-22 and FIPS 140-1 statistical suites, confirming their cryptographic suitability. The encryption follows a confusion–diffusion structure: audio samples are first permuted using chaotic indices and then masked using bitwise XOR with PRNG output. The method is tested on real-world audio types—including speech, environmental sounds, and music—and evaluated through waveform, spectrogram, histogram, and statistical metrics. Results demonstrate strong resistance to brute-force and differential attacks. In addition, the use of a memristor-based system introduces a new perspective in audio security by enabling low-power, hardware-oriented, and physically realizable chaotic circuits. This work contributes a lightweight and secure audio encryption framework suitable for embedded and next-generation multimedia applications.