Secret Communication Systems Using Chaotic Wave Equations with Neural Network Boundary Conditions

Area of Science:

• Applied Mathematics
• Cryptography
• Artificial Intelligence

Background:

• Chaotic synchronization is used in secure communication, embedding information in chaotic signals.
• Previous systems using wave equations with van der Pol boundary conditions had insufficient parameters for robust security.

Purpose of the Study:

• To enhance the security of chaotic synchronization systems for secret communication.
• To develop a more secure method by replacing the nonlinear boundary condition with an artificial neural network.

Main Methods:

• Replaced the nonlinear boundary condition of a wave equation with a two-part artificial neural network.
• Integrated the neural network as left and right boundary conditions for the wave equation.
• Evaluated security performance using monochrome and color images and conducted security tests.

Main Results:

• Encrypted images were nearly identical regardless of the original input, effectively concealing information.
• The Lyapunov exponent indicated that the neural network introduced chaotic vibration.
• Security tests confirmed that transmitted images were encrypted into almost identical wave patterns.

Conclusions:

• The artificial neural network-based boundary condition significantly enhances the security of chaotic synchronization systems.
• The proposed method effectively prevents information retrieval from original images by encrypting them into indistinguishable wave patterns.
• The system demonstrates robust security performance suitable for secret communication applications.