Performance Evaluation of Voice Encryption Techniques Based on Modified Chaotic Systems

With the substantial development of communication systems, especially in the public channels through which the knowledge moves, channel problems are growing. The protection of privacy and data is critical and should be improved, and speech scrambling techniques are one branch of IT that has become an important topic in modern life, primarily used in military security applications. In addition to its applications in the civil side. Speech Scrambling is described as altering speech signals to make them incomprehensible to listeners who are not expected to receive the communication. It will therefore make it difficult for eavesdrops or attackers to understand it. The main objective of this paper is to model and simulate speech scrambling's proposed structure models based on chaotic signals. The scrambling method structure consists of three main components: scrambling, noisy channel and descrambling. The scrambling and descrambling processes have an inverse relationship. On the basis of chaotic signals, the Speech Scrambling was examined using Arabic and English Analog speech signals, spoken by a man or woman of different ages. The paper suggests a voice scrambling models which uses four various modified chaotic system: Random Logistic maps RLM, Random Lorenz system RLS, Random Chen system RCS and Random Unified system RUS based on OFDM technique. The performance evaluation measurements contain five different tests: Mean Square Error MSE, Segmental Spectral Signal to Noise Ratio SSSNR, Entropy H, Correlation Coefficient Rc and Bit Error Rate BER. The objective test to measuring a residual intelligibilities R.I. of voice between the original and scrambled signals for proposed models are: $\mathrm{SSSNR} =-19.5,\ \mathrm{Rc}=0.11,\ \mathrm{MSE} = 0.43$ and $\mathrm{H}=97{\%}$, while the BER between the original signal and recovered signal at 16 dB about 10⁻⁰³. This result is good compared to other results and according to the references that mentioned in the research. Conducted a series of randomness test for chaotic signals as national institute of standards and technology NIST test.