Fast Synchronization Control and Application for Encryption-Decryption of Coupled Neural Networks with Intermittent Random Disturbance

Mathematics Computer Science Engineering Electrical Engineering Information Science Artificial Intelligence Automation
fast synchronization control intermittent random disturbance coupled neural networks encryption-decryption exponential synchronization preset-time synchronization

Fast Synchronization Control and Application for Encryption-Decryption of Coupled Neural Networks With Intermittent Random Disturbance

I. Background and Research Motivation

In recent years, neural networks have been widely applied in various fields such as data classification, image recognition, and combinatorial optimization problems. Regarding the structure and performance of neural networks, they can be divided into deterministic and stochastic neural networks. Many studies have shown that stochastic neural networks with added noise perturbations exhibit better dynamic characteristics than deterministic neural networks. By constructing networks with random perturbations, actual neural network models can be more realistically simulated. However, most current neural network research focuses on continuous perturbation models, although intermittent random disturbances are more common in real-life scenarios.

II. Source of the Paper

This paper, titled “Fast synchronization control and application for encryption-decryption of coupled neural networks with intermittent random disturbance,” was published in the May 2024 issue of the journal “Neural Networks.” The authors include Xianghui Zhou, Jinde Cao, Zhi-Hong Guan, Xin Wang, and Fanchao Kong, who are affiliated with Anhui Normal University, Southeast University, Al-Arabiya University, Huazhong University of Science and Technology, and Huaiyin Normal University, respectively.

III. Research Process and Methods

a) Research Process

The paper designs a new control method for coupled neural networks under intermittent random disturbance and studies fast synchronization control strategies. The design of the controller is based on the Laplacian matrix and some inequality techniques, achieving fast synchronization conditions through Lyapunov stability theory.

  1. Establishing the Neural Network Model: A new coupled neural network model is created, distinguished from other existing stochastic neural networks by adopting intermittent random disturbance noise.

  2. Designing the Controller: Two types of controllers are designed: one with a coupling signal for studying exponential synchronization, and another that not only has an adjustable synchronization rate but also avoids the problem of infinite gain, aimed at research on preset-time synchronization.

  3. Mathematical Model Calculations: Using Lyapunov stability theory, the Laplacian matrix, and some inequality techniques, fast synchronization conditions were obtained.

  4. Numerical Simulation: Numerical examples were used to demonstrate the effectiveness of the control scheme and discuss the impact of different control factors on the synchronization rate.

  5. Application Study: The image encryption and decryption application based on the drive-response network was successfully applied to actual cases.

b) Experimental Results

Numerical examples verified the effectiveness of the designed control scheme and demonstrated exponential synchronization and preset-time synchronization achieved under intermittent random disturbance conditions. Specific results are as follows:

  1. Fast Synchronization Control: Based on the designed controller, conditions for exponential synchronization were derived, proving that neural networks can quickly achieve synchronization under intermittent interference conditions.

  2. Preset-time Synchronization Control: By further optimizing the controller, conditions for achieving synchronization within a preset time were explored, successfully achieving synchronization control within a given time.

  3. Experimental Simulation: Numerical simulations showed the synchronization effects of the designed controller under different parameter conditions, verifying the feasibility of the theoretical conditions proposed in the paper.

c) Research Conclusions

This study proposes a new mode of stochastic disturbance, namely intermittent random disturbance, which better reflects the disturbance conditions in actual networks compared to continuous disturbances. By designing controllers with adjustable synchronization rates, exponential synchronization and preset-time synchronization were achieved. The research results not only theoretically validated the effectiveness but also demonstrated application value in practical applications such as image encryption and decryption.

IV. Research Highlights

  1. Innovative Disturbance Model: Proposes an intermittent random disturbance mode, which can better simulate actual scenarios compared to traditional continuous disturbances.

  2. Fast Synchronization Control: Designed controllers with adjustable synchronization rates, achieving fast synchronization through the Laplacian matrix and Lyapunov stability theory, significantly improving control efficiency and speed.

  3. Practical Application Value: The research results were successfully applied to image encryption and decryption, showing potential applications in secure communication, image processing, and other fields.

  4. Comprehensive Mathematical Validation: Through rigorous mathematical calculations and numerical simulations, the effectiveness and robustness of the proposed methods were validated.

V. Other Valuable Information

  1. Comparison with Other Research: The paper compares the proposed method to other literature in terms of synchronization rate, synchronization mode, and disturbance mode, demonstrating the unique advantages of the method in synchronization control.

  2. Technical Superiority: The proposed intermittent random disturbance model and its synchronization method show advantages in handling complex network environments, providing new theoretical bases and practical tools.

VI. Summary

This paper proposes a new mode of stochastic disturbance and corresponding fast synchronization control methods, achieving significant theoretical breakthroughs and demonstrating high application value in practical scenarios. Through rigorous mathematical validation and actual case studies, the effectiveness and robustness in complex network environments were proven. The research results provide new ideas for synchronization control of neural networks and promote their application in secure communication fields such as image encryption and decryption.