Performance Evaluation of Audio Encryption and Scrambling Algorithms

With the rapid growth of digital communication networks, multimedia data security has become a critical and inescapable requirement for applications such as secure voice communication, military systems, teleconferencing, and telephone banking. Audio signals inherently contain massive data sizes and exhibit a high degree of strong sample correlation, rendering them highly vulnerable to statistical attacks and unauthorized eavesdropping. Conventional block cipher encryption techniques often require extensive computational power and consume excessive energy, making them fundamentally unsuitable for real-time, large-scale audio applications. To address these profound limitations, this study presents a comprehensive and rigorous performance evaluation of various audio security algorithms, specifically focusing on Audio Scrambling, Chaos-based Encryption, and AES-style multi-round Encryption. Evaluated on both synthetic and real-time continuous audio signals, the algorithms are analysed using stringent cryptographic security metrics including the Number of Pixels Change Rate (NPCR), Unified Average Changing Intensity (UACI), Information Entropy, and adjacent sample Correlation Coefficients. The mathematical and experimental results demonstrate that while time-domain scrambling provides basic structural security with minimal computational overhead, chaos-based techniques offer a vastly superior balance, achieving high security and diffusion with optimally low processing times.