The invention discloses a GAN game confrontation
encryption system (method) based on a
chaotic model, relates to a
chaotic encryption method and a confrontation neural network in
machine learning, andgenerates a dynamic
encryption algorithm and an encryption key based on the thought of the game theory. The invention aims to solve the problem that encrypted data is easy to
attack in the existing encryption technology. According to the technical scheme, the
system comprises the following steps: 1, analyzing the characteristics of an
insect mouth logistic
chaotic model; 2, inputting a determined[mu]m value, and generating a chaotic model under the mu value and a one-time pad security key by utilizing
generative adversarial network adversarial; 3, testing
random number generation: testing whether the randomness of the generated sequence is the same as or similar to that of the original sequence or not; 4, setting different loss convergence values, and synthesizing the different loss convergence values and [mu]m into a decision (mu, l) as decision parameters of the
discriminator; and 5, inputting different [mu]m values in an interval where [mu]m is located when the model is in chaos,and generating a
chaos model corresponding to the [mu]m values by utilizing the generation model; and judging when to stop according to the decision parameter, wherein the model obtained when to stopis the GAN encryption
algorithm based on chaos. The invention is applied to the field of communication.