Hybrid audio encryption method based on DCT and Walsh-Hadamard transform using Henon chaotic maps and logistic maps

Abstrakt

A new hybrid audio signal encryption algorithm
is proposed, combining transform-domain processing using the
discrete cosine transform (DCT) and the Walsh–Hadamard
transform (WHT) with the application of Henon chaotic maps
and the logistic map. The novelty of the method lies in the
integration of DCT–WHT transforms to enhance decorrelation
and energy compaction, as well as in an adaptive key selfgeneration
mechanism based on the statistical characteristics
of each audio frame, which eliminates the need for transmitting
and storing external secret data. The algorithm reduces
the correlation between the original and encrypted signals to
approximately 0.0006 (compared to approximately 0.00045 for
the H´enon+LWT method), increases entropy to 5.18 bits (an
improvement of approximately 8.3%), provides NPCR approximately
97.63% (versus 98.47%) and UACI approximately 6.67%
(versus 15.57%), with RMSE of reconstruction 0.0205 (versus
0.0354). Experiments were conducted on the VCTK corpus (16
kHz, 16-bit, 1000 audio segments). In addition to describing the
proposed method, the paper also contains a detailed comparative
analysis of its performance relative to other approaches, including
DCT+H´enon, WHT+H´enon, and H´enon+LWT algorithms. Such
a comparison allows an objective assessment of the advantages
and limitations of each method in the context of cryptographic
security, reconstruction quality, and computational load.