Fractional-order Delayed Hopfield Neural Network Chaotic Source And Its Secure Communication

Chaotic sequences are widely used in information security and secure communication fields due to their inherent complex dynamics such as quasi-randomness and initial value sensitivity.The integer-order low-dimensional chaotic system has security risks,and the fractional-order time-delay system is difficult to be attacked by conventional attack methods and has a larger key space.It has wider application space and practical value in the field of secure communication,so constructing the chaotic system with order delay is an effective way to improve the security of secure communication systems.Hopfield neural network can directly generate chaotic matrix with good diffusion effect under certain conditions,this thesis uses a class of fractional-order time Hopfield neural network as a model to study its complex dynamic characteristics including chaotic phenomena and synchronization problems.It is used in secure communication schemes.Firstly,a new class of fractional-order Hopfield neural networks is proposed and its complex dynamic behavior is analyzed.Theoretically proved the uniqueness of the system equilibrium point.Under the no-delay,the relationship between the fractional order and the stability of the system is given by the fractional stability theorem.When the delay parameters change,the road leading to chaos is found to be a paroxysm chaotic road.,the bifurcation diagram,the largest Lyapunov exponent,and the permutation entropy are used to verify it.Then,based on state observer theory,the generalized projection synchronization problem for a class of fractional-order delay neural networks is studied.The design method of feedback gain matrix is given based on fractional stability theorem and pole placement technique.The numerical simulation experiment is carried out by taking Hopfield neural network with fractional delay as an example.The feasibility and correctness of the synchronization scheme are verified.Finally,based on the state observer synchronization scheme,a chaotic masking secret communication framework is presented.The source signal is directly involved in the generation of chaotic signals after preprocessing.The feasibility of the framework is verified by theoretical analysis and numerical simulation experiments.The synchronization and the encryption and decryption process under the feedback gain matrix parameter mismatch are analyzed.The complexity comparison between our chaotic system and other chaotic systems for encryption is given.Finally,the advantages of the framework over the traditional chaotic cover communication scheme are pointed out.