Speech scrambler with multiwavelet, Arnold Transform and particle swarm optimization

Abstract

Speech scrambling aims to distort speech signals to prevent unauthorized listeners from understanding them, but conventional techniques are vulnerable to attacks. Therefore, more robust and secure speech scrambling algorithms are needed to ensure sensitive communication security. A proposed scheme uses a particle swarm optimization algorithm to generate a random key and optimize the level of noise in the scrambled signal, along with two transformations Multiwavelet and Arnold techniques to improve complexity and security. The proposed algorithm has been evaluated using various performance measurements and has demonstrated superior encryption performance than other similar audio encryption schemes with key space equal to 128 × 2.718. Further research and development in speech scrambling are essential to guarantee secure communication in sensitive contexts such as military and intelligence.