CESARAE: computationally efficient and statistical attack resistant audio encryption system

Audio piracy in the real-time transmission has intensified with digital replication and distribution, making unauthorized sharing effortless and widespread, leading to a significant hurdle in securing the original audio data. To counter this, computationally efficient and robust audio security measures like audio encryption play a pivotal role. In this paper, we suggest CESARAE, a computationally efficient, and robust against statistical attacks audio encryption system based on a new chaotic map called the Logistic Feedback Chaotic Map (LFCM). Firstly, we have developed and tested a new chaotic map LFCM, by adding a controlled feedback coefficient to a conventional logistic map. LFCM has been tested in terms of the Lyapunov exponent, bifurcation diagram, time series plot, phase trajectory, frequency assessment and NIST analysis and it has been found to have improved dynamic behavior and chaotic range making it applicable for real-time applications. Subsequently, LFCM has been used for the development of an audio encryption system, CESARAE. The proposed encryption system works in three stages. The original audio \({A}_{org}\) $A_{\mathit{org}}$ to be encrypted is transformed into a grayscale image \({I}_{audio}\) $I_{\mathit{audio}}$ in the first stage. The auditory image is substituted in the subsequent stage, and a concurrent substitution and permutation procedure is carried out in the final step to obtain encrypted audio. CESARAE has been found to meet security goals, yielding average metrics of Correlation Coefficient CC = 0.0032, Number of Sample Change Rate NSCR = 99.62%, Unified Average Change Intensity UACI = 33.53%, Root Mean Square RMS( \({A}_{org})\) $A_{\mathit{org}})$ = 0.111, Crest Factor CF( \({A}_{org}\) $A_{\mathit{org}}$ ) = 13.7322, Root Mean Square RMS( \({Audio}_{encrypt}\) ${\mathit{Audio}}_{\mathit{encrypt}}$ ) = 0.3001, CF( \({Audio}_{encrypt}\) ${\mathit{Audio}}_{\mathit{encrypt}}$ ) = 5.3737, Signal-to-Noise Ratio SNR = 8.4035, Peak Signal-to-Noise Ratio PSNR ( \({A}_{org}, {Audio}_{encrypt}\) $A_{\mathit{org}},{\mathit{Audio}}_{\mathit{encrypt}}$ ) = 4.0816, and Peak Signal-to-Noise Ratio PSNR ( \({A}_{org}\) $A_{\mathit{org}}$ , \({Audio}_{decrypt}\) ${\mathit{Audio}}_{\mathit{decrypt}}$ ) = 59.3631,coupled with an expansive key space of \({2}^{300.207}\) $2^{300.207}$ —far surpassing existing methods. Empirical outcomes underscore its robust security and satisfactory encryption/decryption speeds, rendering it a viable solution for secure communication and real-time data transfer.