Neural Network Prediction Method of Chaotic FH Code Sequence

摘要

The frequency hopping communication implements the communication in the process of continuous and irregular jumping frequency, widely used in military communications and civilian communications. At present, the research on the modeling and prediction of frequency hopping codes by using chaos analysis and neural network and so on algorithms is very closely developed and has achieved some effects. In this paper, the RBF neural network was used to conduct simulation prediction of the m code, the RS code, and the nonlinear code these three frequency hopping codes, and the simulation experiments were carried out by MATLAB. The performance of the prediction model was analyzed and compared by theoretical analysis and simulation results. The results showed that the RBF neural network was more powerful in approximation ability, classification learning, learning speed and so on aspects.