Chaotic masking of voice bitstream using mixed sequences for authorized listening

Abstract

The voice-based communication becomes extensively vital in the application areas of military, voice over IP, voice-conferencing, phone banking, news telecasting etc. It greatly demands to preserve sensitive voice signals from the unauthorized listening and illegal usage over shared/open networks. To address the need, we propose a chaos-based symmetric encryption technique to protect voice bitstreams over insecure transmission channel. The technique utilizes the features of high dimensional chaos like Lorenz and Chen systems to generate highly unpredictable and random-like sequences. The encryption keys are dynamically extracted from the pretreated chaotic mixed sequences, which are then used to mask the voice bitstream for integrity protection of voice data. The experimental analyses like auto-correlation, signal distribution, parameter-residual deviation, key space and key-sensitivity demonstrate the effectiveness of the proposed technique for secure voice communication. © Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2012.